Nationwide population-based study on the use of novel antiandrogens in men with prostate cancer in Sweden

Figures & data

Table 1. Characteristics of men in Prostate Cancer data Base Sweden (PCBaSe) RAPID with prostate cancer registered between 1 January 1998 and 31 December 2014, in the National Prostate Cancer Register and residing in Sweden on 1 July 2015.

	NOVA (%)	Abiraterone (%)	Enzalutamide (%)	All men (%)
Men	1650 (100)	518 (100)	1210 (100)	91,209 (100)
Calendar year of diagnosis
1998–2005	450 (27)	155 (30)	311 (26)	23,613 (26)
2006–2008	440 (27)	150 (29)	320 (26)	25,079 (27)
2009–2012	438 (27)	128 (25)	326 (27)	23,431 (26)
2013–2014	322 (20)	85 (16)	253 (21)	19,086 (21)
Time between diagnosis and NOVA (months)
Median (IQR)	75 (40–124)	79 (42–126)	72 (39–120)
<18	85 (5)	22 (4)	70 (6)
18–59	573 (35)	167 (32)	428 (35)
60–119	560 (34)	186 (36)	407 (34)
≥120	432 (26)	143 (28)	305 (25)
Age at diagnosis (years)
Median (IQR)	68 (63–73)	67 (62–73)	68 (63–73)	67 (62–72)
<60	218 (13)	73 (14)	155 (13)	15,595 (17)
60–69	732 (44)	244 (47)	534 (44)	43,283 (47)
70–79	600 (36)	170 (33)	450 (37)	26,207 (29)
80–89	99 (6)	31 (6)	70 (6)	5864 (6)
≥90	1 (0)	0 (0)	1 (0)	260 (0)
Age at NOVA (years)
Median (IQR)	75 (70–80)	75 (69–80)	74 (70–80)
<60	51 (3)	16 (3)	37 (3)
60–69	360 (22)	120 (23)	262 (22)
70–79	783 (47)	243 (47)	582 (48)
80–89	444 (27)	135 (26)	321 (27)
≥90	12 (1)	4 (1)	8 (1)
CCI on 1 July 2015
0	1012 (61)	321 (62)	739 (61)	58,041 (64)
1	253 (15)	77 (15)	188 (16)	14,133 (15)
≥2	385 (23)	120 (23)	283 (23)	19,035 (21)
Marital status
Never married	175 (11)	44 (8)	140 (12)	9260 (10)
Divorced/separated/widowed	311 (19)	93 (18)	234 (19)	18,385 (20)
Married	1164 (71)	381 (74)	836 (69)	63,555 (70)
Missing data	0 (0)	0 (0)	0 (0)	9 (0)
Educational level^a
Low	572 (35)	174 (34)	427 (35)	29,649 (33)
Middle	631 (38)	189 (36)	472 (39)	36,359 (40)
High	433 (26)	151 (29)	301 (25)	24,667 (27)
Missing data	14 (1)	4 (1)	10 (1)	534 (1)
Income in quartiles^b
Q1	454 (28)	126 (24)	344 (28)	22,801 (25)
Q2	476 (29)	142 (27)	357 (30)	22,801 (25)
Q3	381 (23)	129 (25)	273 (23)	22,800 (25)
Q4	339 (21)	121 (23)	236 (20)	22,801 (25)
Missing data	0 (0)	0 (0)	0 (0)	6 (0)
Diagnosed in a county with a university hospital
No	780 (47)	264 (51)	567 (47)	36,960 (41)
Yes	870 (53)	254 (49)	643 (53)	54,249 (59)
Risk category^c
Favorable-risk	294 (18)	97 (19)	206 (17)	62,368 (68)
Aggressive disease	1350 (82)	419 (81)	1000 (83)	26,631 (29)
Missing data	6 (0)	2 (0)	4 (0)	2210 (2)
Primary treatment
Conservative management	74 (5)	18 (4)	58 (5)	21,386 (26)
Radical prostatectomy	195 (13)	72 (15)	134 (12)	31,996 (38)
Radiotherapy with curative intent	217 (14)	64 (13)	166 (15)	16,637 (20)
Antiandrogens	220 (15)	70 (15)	161 (15)	4652 (6)
GnRH agonists	799 (53)	252 (53)	583 (53)	8941 (11)

Data are shown as n (%) or median (interquartile range).

Filling of prescriptions for a novel antiandrogen (NOVA) was determined between 1 July 2015 and 30 April 2016.

a Low = compulsory school, <10 years; middle = upper secondary school, 10–12 years; high = college or university, >12 years;

b based on men with prostate cancer living in Sweden on 1 July 2015;

c favorable-risk includes low-risk and intermediate-risk cancer; aggressive disease includes high-risk, locally advanced, regionally metastatic and distant metastases. Prostate cancer risk categories: low-risk = T1–2, prostate-specific antigen (PSA) < 10 ng/ml and Gleason score (GS) ≤ 6; intermediate-risk = T1–2, PSA <20 ng/ml and PSA 10–19.9 ng/ml or GS 7; high-risk = T1–2, PSA <50 ng/ml with GS 8–10 or PSA 20–49.9 ng/ml; locally advanced = T3 and PSA <50 ng/ml; regionally metastatic = PSA <100 ng/ml with T4 or PSA 50–99.9 ng/ml or N1; distant metastases = PSA ≥100 ng/ml or M1.

CCI: Charlson Comorbidity Index; GnRH: gonadotropin-releasing hormone.

Figure 1. Timeline of the study period in the Prostate Cancer data Base Sweden (PCBaSe) RAPID. NPCR: National Prostate Cancer Register; PDR: Prescribed Drug Register; NOVA: novel antiandrogen; FU: follow-up; TPR: Total Population Register (vital status). The grey rectangle represents the study period for filled prescriptions for NOVAs in the PDR. A–E illustrate the different periods during which a patient could be at risk before death or censoring: patient A was at risk during the whole study period; patients B and C died during the study period; patient D died before 1 November 2015 and is not included in the analysis of users vs non-users; patient E died before 1 July 2015 and is not included in the study.

Table 2. Characteristics of men with a high probability of castration-resistant prostate cancer in the Prostate Cancer data Base Sweden (PCBaSe) RAPID, registered in the National Prostate Cancer Register between 1 January 1998 and 31 December 2014 with distant metastases at the date of diagnosis, on gonadotropin-releasing hormone (GnRH) therapy on 31 December 2014 and residing in Sweden on 1 November 2015.

	Received NOVA (%)	Did not receive NOVA (%)
Men	421 (100)	1493 (100)
Calendar year of diagnosis
1998–2005	21 (5)	131 (9)
2006–2008	61 (14)	226 (15)
2009–2012	144 (34)	441 (30)
2013–2014	195 (46)	695 (47)
Time between diagnosis and NOVA (months)
Median (IQR)	36 (23–58)
<18	61 (14)
18–59	258 (61)
60–119	85 (20)
≥120	17 (4)
Age at NOVA (years)
Median (IQR)	72 (67–78)
< 60	21 (5)
60–69	126 (30)
70–79	203 (48)
80–89	69 (16)
≥90	2 (0)
CCI on 1 July 2015
0	257 (61)	855 (57)
1	57 (14)	247 (17)
≥2	107 (25)	391 (26)
Marital status
Never married	58 (14)	170 (11)
Divorced/separated/widowed	76 (18)	349 (23)
Married	287 (68)	974 (65)
Missing data	0 (0)	0 (0)
Educational level^a
Low	134 (32)	604 (40)
Middle	155 (37)	582 (39)
High	125 (30)	292 (20)
Missing data	7 (2)	15 (1)
Income in quartiles^b
Q1	93 (22)	456 (31)
Q2	126 (30)	485 (32)
Q3	96 (23)	318 (21)
Q4	106 (25)	234 (16)
Missing data	0 (0)	0 (0)
Diagnosed in a county with a university hospital
No	203 (48)	698 (47)
Yes	218 (52)	795 (53)
Time between GnRH and NOVA (months)
<18	82 (19)
18–59	250 (59)
60–119	76 (18)
≥120	13 (3)

Data are shown as n (%) or median (interquartile range).

Filling of prescriptions for a novel antiandrogens (NOVA), abiraterone or enzalutamide, was determined between 1 July 2015 and 15 April 2016.

a Low = compulsory school, <10 years; middle = upper secondary school, 10–12 years; high = college or university, >12 years;

b based on men with prostate cancer living in Sweden on 1 July 2015.

CCI: Charlson Comorbidity Index.

Table 3. Odds ratios (ORs) and 95% confidence intervals (CIs) for receipt of a novel antiandrogen (NOVA).

			Univariable		Multivariable
	No. at risk	No. of events	OR	95% CI	OR	95% CI
Age at diagnosis (years)
<70	789	227	1.00	(Ref.)	1.00	(Ref.)
70–79	742	163	0.70	(0.55–0.88)	0.75	(0.59–0.96)
≥ 80	383	31	0.22	(0.15–0.32)	0.23	(0.15–0.35)
CCI on 1 July 2015
0	1112	257	1.00	(Ref.)	1.00	(Ref.)
1	304	57	0.77	(0.56–1.06)	0.92	(0.66–1.28)
≥2	498	107	0.91	(0.71–1.18)	1.11	(0.85–1.44)
Marital status
Never married	228	58	1.00	(Ref.)	1.00	(Ref.)
Divorced/separated/widowed	425	76	0.64	(0.43–0.94)	0.76	(0.51–1.14)
Married	1261	287	0.86	(0.62–1.20)	0.92	(0.66–1.29)
Educational level^a
Low	738	134	1.00	(Ref.)	1.00	(Ref.)
Middle	737	155	1.20	(0.93–1.55)	1.10	(0.85–1.44)
High	417	125	1.93	(1.46–2.56)	1.64	(1.23–2.20)
Diagnosed in a county with a university hospital
No	901	203	1.00	(Ref.)	1.00	(Ref.)
Yes	1013	218	0.94	(0.76–1.17)	0.93	(0.74–1.16)

Logistic regression model including men registered in the National Prostate Cancer Register between 1 January 1998 and 31 December 2014 with metastatic cancer, on gonadotropin-releasing hormone (GnRH) therapy on 31 December 2014. All men included in the table were residing in Sweden on 1 November 2015. The outcome, filling of prescriptions for a NOVA, abiraterone or enzalutamide, is determined between 1 July 2015 and 15 April 2016.

a Low: compulsory school, < 10 years; middle: upper secondary school, 10–12 years; high: college or university, >12 years.

CCI: Charlson Comorbidity Index.

Figure 2. Proportion of men with prostate cancer (PCa) diagnosed in 2014 who had a prescription filled for (a) a novel antiandrogen (NOVA), (b) abiraterone or (c) enzalutamide, between 1 July 2015 and 30 April 2016 per 1000 deaths from PCa in each county in 2013–2015.

Figure 3. Odds ratios (ORs) and 95% confidence intervals (CIs) for receipt of novel antiandrogens (NOVAs) according to county of residence: logistic regression model including men registered in the National Prostate Cancer Register (NPCR) between 1 January 1998 and 31 December 2014 with metastatic cancer, on gonadotropin-releasing hormone (GnRH) agonist on 31 December 2014. All men included in the figure were at risk on 1 November 2015. Filling of a prescription for a novel antiandrogen (NOVA), abiraterone or enzalutamide, was determined between 1 July 2015 and 15 April 2016. A logarithmic scale (base 10) was used. The adjusted model includes year of diagnosis, age, comorbidity, marital status and educational level.