Abstract
Background:
PDE5I use has been linked to a number of ocular side effects, such as serous retinal detachment (SRD), retinal vascular occlusion (RVO), and ischemic optic neuropathy (ION).
Aim:
We investigated the risk for SRD, RVO, and ION in patients using PDE5Is.
Methods:
We utilized the IBM® MarketScan® (2007 to 2021) Commercial and Medicare Supplemental Databases (v 2.0) for this analysis. To estimate overall events risk, Cox proportional hazard models were applied to calculate the hazard ratios for ED diagnosis and the different treatments adjusting for region, median age, obesity, DM, hyperlipidemia, smoking, HTN, CAD, and sleep apnea. Additionally, the same analyses were performed to calculate the hazard ratios for BPH diagnosis and the different treatments.
Outcomes:
Hazard Ratio for SRD, RVO, and ION.
Results:
In total, 1,938,262 men with an ED diagnosis were observed during the study period. Among them, 615,838 (31.8%) were treated with PDE5i. In total, 2,175,439 men with a BPH diagnosis were observed during the study period. Among them, 175,725 (8.1%) were treated with PDE5i. On adjusted Cox regression analysis, PDE5i use was not associated with SRD, RVO, ION, and any ocular event when compared to ED diagnosis and other ED treatments. Importantly, as the intensity of ED treatment increased, so did the risk of ocular events. In addition, PDE5i use was not associated with SRD and ION when compared to BPH diagnosisis and other BPH treatments. In contrast, in BPH patients PDE5i use was associated with RVO (HR: 1.14; 95% CI: 1.06 −1.23). Importantly, BPH patients receiving other medical treatment (i.e. 5a reductase/alpha blocker) (HR: 1.11; 95% CI: 1.06–1.16) or surgical treatment (HR: 1.10; 95% CI: 1.02–1.19) had a higher risk of RVO.
Clinical Implications:
We did not observe any consistent association between PDE5i use and any ocular adverse events (SRD, RVO, and ION).
Strengths and Limitations:
Because we did not have access to the patient’s medical records, the outcome definitions were recorded using ICD-9 and ICD-10 coding.
Conclusions:
Patients using PDE5i for ED or BPH indications did not have an increased risk of ocular events even when compared to other ED treatments or BPH treatments.
Keywords: erectile dysfunction, pde5i, adverse events
1. Introduction
A disturbance in any of the physiological, relational, or psychological components of the erectile response results in erectile dysfunction (ED), a multifaceted but frequent male sexual dysfunction1. A man’s quality of life can be negatively impacted by erectile dysfunction, and the majority of patients experience depression and anxiety symptoms that are related to sexual performance2.
Over the past 20 years, strong data has accumulated showing how several modifiable and immutable risk factors are connected to the pathophysiology of ED3–6. As a result, the cornerstone of ED treatment has steadily shifted toward a more individualized and customized approach, in which the patient and the specialist discuss the benefits and risks of a particular therapeutic pathway and jointly choose a specific and desired outcome7,8. Due to their positive impact on erectile function and favorable tolerability, phosphodiesterase-5 inhibitors (PDE5i) are often the first-line therapeutic option in males with ED9, and PDE5Is were prescribed over 20 million times per month in the US in 202010.
While visual disturbances (e.g. cyanopsia) may occur with PDE5i use, recent data also reported a number of serious ocular side effects, such as serous retinal detachment (SRD), retinal vascular occlusion (RVO), and ischemic optic neuropathy (ION)11,12. However, there is a significant overlap between the risk factors for major ocular events and ED2,13. Indeed, the most frequent cause of erectile dysfunction, just as SRD, RVO, and ION are vascular and metabolic disease14. We thus aimed to investigate the risk for SRD, RVO, and ION in patients using PDE5Is and compare the rates of SRD, RVO, and ION in other men with diseases in which PDE5i are prescribed (i.e. ED and benign prostatic hyperplasia (BPH)).
2. Methods
2.1. Patients
We utilized the IBM® MarketScan® (2007 to 2021) Commercial and Medicare Supplemental Databases (v 2.0) for this analysis. These databases contain de-identified and individual-level healthcare claims information from employers, health plans, hospitals, and Medicare and Medicaid programs. They provide reimbursed health care claims data on inpatient, outpatient, and pharmacy encounters covering over 150 million individuals who are privately insured through employment-sponsored or Medicare supplementary health insurance. These datasets are fully compliant with U.S. privacy laws and regulations. Institutional review board approval was not required for the analysis because this data set contains de-identified patient information.
Men aged 18+ years were identified from the inpatient and outpatient files. Men aged <18 years, those who had an ocular event before the diagnosis of ED or BPH and those who were treated before receiving a diagnosis of ED or BPH were excluded. Based on the common approved treatment indications for PDE, we identified two cohorts: 1) men diagnosed with ED (ICD-9:302.70, 302.72, 607.84, 608.89; ICD-10: R37, N52, F52.21, N53.9 and 2) those diagnosed with benign prostatic hyperplasia (BPH) (ICD-10: N40, ICD-9: 600.00, 600.01, 600.20, 600.21, 600.90, 600.91). We recorded the first date of a relevant diagnosis as the index date, and grouped men with a diagnosis code of ED into 4 categories: (1) men diagnosed with ED and not treated; (2) men diagnosed with ED treated with intracavernosal injection (ICI) or urethral suppository treatment; (3) men diagnosed with ED treated with PDE5i; (4) men diagnosed with ED treated with penile prosthesis (PP). Similarly, we grouped men with a diagnosis code of BPH into 4 categories: (1) men diagnosed with BPH and not treated; (2) men diagnosed with BPH treated with 5a reductase or alpha-blocker; (3); men diagnosed with BPH treated with PDE5i; (4) men diagnosed with BPH treated surgically (CPT: 52601, 52620, 52630, 52648, 62649, 55801, 0421T, 54405, 54408, 54410, 54400, 54416, 54235, J0275).
2.2. Outcome Ascertainment
Outcomes were defined as new cases of SRD (ICD-9: 361.2, 362.4, 362.41, 362.42; ICD-10: H35.711, H35.712, H35.713, H33.23, H33.21, H33.22), RVO (ICD-9: 362.30, 362.31, 362.35; ICD-10: H34), or ION (ICD-9: 377.41; ICD-10: H47.01, H47.011, H47.012)11. We performed three separate analyses and also identified cases as a composite of any of the three outcomes.
2.3. Statistical Analysis
Frequencies (percentages) were reported for categorical variables, means (standard deviation) or medians (range) were reported for continuous variables by the exposure groups. To compare covariates between the exposure groups, the Chi-square test was implemented for categorical variables, s. To estimate overall events risk, Cox proportional hazard models were applied to calculate the hazard ratios for ED diagnosis and the different treatments adjusting for region, median age, obesity, DM, hyperlipidemia, smoking, HTN, CAD, and sleep apnea. Time to the outcome events was calculated from their index dates until the first occurrence of SRD, RVO, or ION, or censored at the last enrollment date. Additionally, the same analyses were performed to calculate the hazard ratios for BPH diagnosis and the different treatments. The interaction terms with median age, comorbidity index categories, and the number of drug units dispensed of each group was checked and the hazard ratios at different levels were calculated to identify their differences. All P values were 2-sided with P < 0.05 considered statistically significant. Sensitivity analyses were performed look at men above vs below the median age for the ED and BPH cohorts. In addition, cohorts were stratified based on patient health as assessed by the Charlson Comorbidity Index15,16. Analyses were performed using SAS (version 9.4, SAS Institute, Inc., Cary, NC).
3. Results
In total, 1,938,262 men with an ED diagnosis were observed during the study period. Among them, 615,838 (31.8%), 14,780 (0.76%), 9,313 (0.48%), and 1,298,331 (67.0%) were treated with PDE5i, ICI/urethral suppository, PP, and not treated, respectively. The mean (SD) follow-up was 4.1 (0.004), 4.7 (0.03), 5.2 (0.04), and 3.0 (0.002) years for men treated with PDE5i, ICI/urethral suppository, PP, and men not treated, respectively. In total, 2,175,439 men with a BPH diagnosis were observed during the study period. Among them 175,725 (8.1%), 744,755 (34.2%), 76,724 (3.5%), and 1,178,235 (54.2%) were treated with PDE5i, 5a reductase/alpha blocker, surgically, and not treated, respectively. The mean (SD) follow-up was 4.5 (0.007), 3.9 (0.004), 4.9 (0.012), and 3.1 (0.002) years for men treated with PDE5i, 5a reductase/alpha blocker, surgically treated, and men not treated, respectively. Demographic and clinical characteristics are summarized in Table 1.
Table 1.
Patient’s demographic and clinical characteristics
| Total | ED with no treatment | ED + PP | ED+ICI | ED+PDE5i | Total | BPH with no treatment | BPH+TURP | BPH+PDE5i | BPH + 5a reductase/alpha blocker | |
|---|---|---|---|---|---|---|---|---|---|---|
| N. (%) | 1,938,262 | 1,298,331 (67.0) | 9,313 (0.48) | 14,780 (0.76) | 615,838 (31.8) | 2,175,439 | 1,178,235 (54.2) | 76,724 (3.5) | 175,725 (8.1) | 744,755 (34.2) |
| Avg Follow up, Years | 3.0 | 5.2 | 4.7 | 4.1 | 3.1 | 4.9 | 4.5 | 3.9 | ||
| Demographic | ||||||||||
| Median Age (IQR), Years | 53 [43 – 60] | 61 [55 – 67] | 60 [54 – 65] | 55 [48 – 61] | 58 [52 – 64] | 67 [60 – 76] | 58 [53 – 63] | 62 [56 – 71] | ||
| Age Categories (%), N. | ||||||||||
| <20 | 14,506 | 14,055 (1.1) | <11 | <11 | 440 (0.07) | 567 | 501 (0.04) | 0 (0) | <11 | 60+ |
| 20–29 | 84,976 | 76,922 (5.9) | 20+ | 60+ | 7,963 (1.3) | 7.235 | 5,624 (0.48) | 18 (0.02) | 160+ | 1,425 (0.19) |
| 30–39 | 188,237 | 146,721 (11.3) | 132 (1.4) | 285 (1.9) | 41,099 (6.7) | 40,836 | 28,520 (2.4) | 152 (0.20) | 2,765 (1.6) | 9,399 (1.3) |
| 40–49 | 427,781 | 291,251 (22.4) | 873 (9.4) | 1,522 (10.3) | 134,135 (21.8) | 258,677 | 170,992 (14.5) | 2,110 (2.8) | 23,493 (13.4) | 62,082 (8.3) |
| 50–59 | 668,965 | 420,689 (32.4) | 3,133 (33.6) | 5,340 (36.1) | 239.803 (38.9) | 747,918 | 442,158 (37.5) | 16,424 (21.4) | 72,428 (41.2) | 216,908 (29.1) |
| 60+ | 553,797 | 348,693 (26.9) | 5,143 (55.2) | 7,563 (51.2) | 192,398 (31.2) | 1,120,206 | 530,440 (45.0) | 58,020 (75.6) | 76,866 (43.7) | 454,880 (61.1) |
| Region (%), N. | ||||||||||
| Northeast | 324,963 | 214,685 (16.5) | 1,211 (13.0) | 2,741 (18.6) | 106,326 (17.3) | 444,967 | 283,611 (24.1) | 11,429 (14.9) | 32,404 (18.4) | 117,523 (15.8) |
| North Central | 443,071 | 289,263 (22.3) | 2,779 (29.8) | 3,808 (25.8) | 147,221 (23.9) | 520,409 | 246,564 (20.9) | 24,337 (31.7) | 42,970 (24.5) | 206,538 (27.7) |
| South | 800,535 | 550,198 (42.4) | 4,120 (44.2) | 5,529 (37.4) | 240,688 (39.1) | 802,806 | 418,049 (35.5) | 26,825 (35.0) | 74,039 (42.1) | 283,893 (38.1) |
| West | 333,555 | 216,095 (16.6) | 1,126 (12.1) | 2,535 (17.2) | 113,799 (18.5) | 369,331 | 202,107 (17.2) | 13,007 (17.0) | 24,673 (14.0) | 129,544 (17.4) |
| Unknown | 36,138 | 28,090 (2.2) | 77 (0.83) | 167 (1.1) | 7,804 (1.3) | 37,926 | 27,904 (2.4) | 1,126 (1.5) | 1,639 (0.93) | 7,257 (0.97) |
| Mean N. of Office visits / year | ||||||||||
| 0 | 76,058 | 57,502 (4.4) | 204 (2.2) | 247 (1.7) | 18,105 (2.9) | 69,736 | 44,987 (3.8) | 2,857 (3.7) | 3,075 (1.8) | 18,817 (2.5) |
| 1 | 90,813 | 68,113 (5.3) | 180 (1.9) | 281 (1.9) | 22,239 (3.6) | 80,802 | 54,728 (4.6) | 2,085 (2.7) | 3,793 (2.2) | 20,196 (2.7) |
| 2 | 101,519 | 74,822 (5.8) | 194 (2.1) | 295 (2.0) | 26,208 (4.3) | 92,438 | 61,609 (5.2) | 2,349 (3.1) | 4,677 (2.7) | 23,803 (3.2) |
| 3 | 103,059 | 75,077 (5.8) | 216 (2.3) | 344 (2.3) | 27,422 (4.5) | 97,406 | 63,418 (5.4) | 2,633 (3.4) | 5,253 (3.0) | 26,102 (3.5) |
| 4+ | 1,566,813 | 1,022,817 (78.8) | 8,519 (91.5) | 13,613 (92.1) | 521,864 (84.7) | 1,835,057 | 953,493 (80.9) | 66,800 (87.1) | 158,927 (90.4) | 655,837 (88.1) |
| Smoking | ||||||||||
| No | 1,738,109 | 1,164,234 (89.7) | 8,356 (89.7) | 13,586 (91.9) | 551,933 (89.6) | 1,981,279 | 1,081,164 (91.8) | 70,548 (92.0) | 158,390 (90.1) | 671,177 (90.1) |
| Yes | 200,153 | 134,097 (10.3) | 957 (10.3) | 1,194 (8.1) | 63,905 (10.4) | 194,160 | 97,071 (8.24) | 6,176 (8.1) | 17,335 (9.9) | 73,578 (9.9) |
| Comorbidities | ||||||||||
| Diabetes | ||||||||||
| No | 1.544.903 | 1,049,357 (80.8) | 5,374 (57.7) | 9,403 (63.6) | 480,769 (78.1) | 1.665.115 | 932,146 (79.1) | 55,121 (71.8) | 129,436 (73.7) | 548,412 (73.6) |
| Yes | 393.359 | 248,974 (19.2) | 3,939 (42.3) | 5,377 (36.4) | 135,069 (21.9) | 510.324 | 246,089 (20.9) | 21,603 (28.2) | 46,289 (26.3) | 196,343 (26.4) |
| Obesity | ||||||||||
| No | 1.711.096 | 1,144,593 (88.2) | 8,245 (88.5) | 13,287 (89.9) | 544,971 (88.5) | 1.953.708 | 1,067,702 (90.6) | 70,779 (92.3) | 155,091 (88.3) | 660,136 (88.6) |
| Yes | 227.166 | 153,738 (11.8) | 1,068 (11.5) | 1,493 (10.1) | 70,867 (11.5) | 221.731 | 110,533 (9.4) | 5,945 (7.8) | 20,634 (11.7) | 84,619 (11.4) |
| Hypertension | ||||||||||
| No | 982.406 | 687,394 (2.9) | 2,825 (30.3) | 5,237 (35.4) | 286,950 (46.6) | 908.177 | 540,518 (45.9) | 27,729 (36.1) | 66,241 (37.7) | 273,689 (36.8) |
| Yes | 955.856 | 610,937 (47.1) | 6,488 (69.7) | 9,543 (64.6) | 328,888 (53.4) | 1.267.262 | 637,717 (54.1) | 48,995 (63.9) | 109,484 (62.3) | 471,066 (63.3) |
| Coronary Artery Disease | ||||||||||
| No | 1.728.977 | 1,164,538 (89.7) | 6,960 (74.7) | 11,489 (77.7) | 545,990 (88.7) | 1.777.086 | 997,684 (84.7) | 56,818 (74.1) | 145,747 (82.9) | 576,837 (77.5) |
| Yes | 209.285 | 133,793 (10.3) | 2,353 (25.3) | 3,291 (22.3) | 69,848 (11.3) | 398.353 | 180,551 (15.3) | 19,906 (25.9) | 29,978 (17.1) | 167,918 (22.6) |
| Hyperlipidemia | ||||||||||
| No | 914.862 | 644,165 (49.6) | 3,274 (35.2) | 5,324 (36.0) | 262,099 (42.6) | 841.242 | 468,983 (39.8) | 32,607 (42.5) | 58,886 (33.5) | 280,766 (37.7) |
| Yes | 1.023.400 | 654,166 (50.4) | 6,039 (64.8) | 9,456 (64.0) | 353,739 (57.4) | 1.334.197 | 709,252 (60.2) | 44,117 (57.5) | 116,839 (66.5) | 463,989 (62.3) |
| Sleep Apnea | ||||||||||
| No | 1.809.723 | 1,213,157 (93.4) | 8,478 (91.0) | 13,592 (92.0) | 574,496 (93.3) | 2.030.289 | 1,114,847 (94.6) | 72,018 (93.9) | 159,677 (90.9) | 683,747 (91.8) |
| Yes | 128.539 | 85,174 (6.6) | 835 (9.0) | 1,188 (8.0) | 41,342 (6.7) | 145.150 | 63,388 (5.4) | 4,706 (6.1) | 16,048 (9.1) | 61,008 (8.2) |
| Sum of comorbidities | ||||||||||
| 0 | 542.001 | 396,715 (30.6) | 1,013 (10.9) | 1,993 (13.5) | 142,280 (23.1) | 429.135 | 265,945 (22.6) | 12,898 (16.8) | 27,200 (15.5) | 123,092 (16.5) |
| 1 | 450.008 | 295,713 (22.8) | 1,739 (18.7) | 3,155 (21.4) | 149,401 (24.3) | 506.717 | 285,199 (24.2) | 17,697 (23.1) | 39,872 (22.7) | 163,949 (22.0) |
| 2 | 433.231 | 276,122 (21.3) | 2,399 (25.8) | 3,807 (25.8) | 150,903 (24.5) | 547.843 | 293,943 (25.0) | 19,443 (25.3) | 46,854 (26.7) | 187,603 (25.2) |
| 3+ | 513.022 | 329,781 (25.4) | 4,162 (44.7) | 5,825 (39.4) | 173,254 (28.1) | 691.744 | 333,148 (28.3) | 26,686 (34.8) | 61,799 (35.2) | 270,111 (36.3) |
| Medications, Median [IQR] | ||||||||||
| Total quantity | 240 [60 – 690] | 122 [30 – 630] | 50 [12 – 229] | 48 [15 – 174] | 22 [8 – 58] | 450 [150 – 1170] | 192 [46 – 708] | 390 [120 – 990] | ||
| Total days supply | 270 [60 – 720] | 210 [60 – 726] | 148 [42 – 483] | 103 [30 – 342] | 60 [24 – 152] | 480 [150 – 1220] | 304 [90 – 858] | 390 [120 – 990] | ||
| Total fills | 5 [2 – 13] | 6 [2 – 16] | 5 [2 – 15] | 5 [2 – 12] | 3 [1 – 6] | 10 [4 – 22] | 9 [3 – 20] | 8 [3 – 18] | ||
| Numbers of units dispensed | ||||||||||
| 1–29 | 211.461 | 9,925 (7.1) | 883 (21.7) | 5,199 (39.9) | 195,454 (38.0) | 73.802 | 27,789 (56.8) | 1,509 (2.9) | 25,145 (16.9) | 19,359 (3.0) |
| 30–59 | 106.871 | 20,326 (14.5) | 535 (13.2) | 1,677 (12.9) | 84,333 (16.4) | 94.909 | 9,013 (18.4) | 3,729 (7.1) | 16,623 (11.2) | 65,544 (10.1) |
| 60–119 | 94.844 | 19,328 (13.7) | 545 (13.4) | 1,597 (12.3) | 73,374 (14.3) | 105.562 | 6,586 (13.5) | 5,831 (11.1) | 18,738 (12.6) | 74,407 (11.4) |
| 120–179 | 43.400 | 8,471 (6.0) | 255 (6.3) | 771 (5.9) | 33,903 (6.6) | 52.329 | 2,419 (45.0) | 3,525 (6.7) | 10,467 (7.0) | 35,918 (5.5) |
| 180+ | 215.609 | 82,609 (58.8) | 1,846 (45.4) | 3,782 (29.0) | 127,372 (24.8) | 575.267 | 3,109 (6.4) | 38,057 (72.3) | 77,629 (52.2) | 456,472 (70.0) |
Table 2 summarizes the different rates of ocular events among the different cohorts and the relative hazard ratios (HRs). Among patients diagnosed with ED, 2,972 (0.23%), 3,262 (0.25%), and 1,098 (0.08%) experienced SRD, RVO, and ION, respectively. Among patients diagnosed with ED and treated with PDE5i, 2,079 (0.34%), 2,523 (0.41%), and 751 (0.12%) experienced SRD, RVO, and ION, respectively. Among patients diagnosed with ED and treated with ICI, 84 (0.57%), 120 (0.81%), and 41 (0.28%) experienced SRD, RVO, and ION, respectively. Among patients diagnosed with ED and treated with PP, 43 (0.46%), 103 (1.1%), and 30 (0.32%) experienced SRD, RVO, and ION, respectively. On adjusted Cox regression analysis, PDE5i use was not significantly associated with SRD, ION, and any ocular event when compared to ED diagnosis and other ED treatments. ED patients treated with PP had a higher risk of RVO (HR: 1.24; 95% CI: 1.01–1.50) when compared to untreated ED patients and those treated with other modalities. Figure 1 depicts the HRs for all ED patients populations predicting any ocular event.
Table 2.
Rates and Risks of Ocular Events across all cohorts. SRD: serous retinal detachment; RVO: retinal vascular occlusion; ION: ischemic optic neuropathy; ED: erectile dysfunction; PP: penile prosthesis; ICI: intra-cavernous injection; PDE5i: phosphodiesterase type 5 inihibitors; BPH: benign prostatic hyperplasia; unHR: unadjusted hazard ratio; aHR: hazard ratio adjusted by region, median age, obesity, DM, hyperlipidemia, smoking, HTN, CAD, and sleep apnea.
| N. of patients | SRD | RVO | ION | SRD, RVO, or ION | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| n (%) | unHR (95% CI) | aHR (95% CI) | p | n (%) | unHR (95% CI) | aHR (95% CI) | p | n (%) | unHR (95% CI) | aHR (95% CI) | p | n (%) | uHR (95% CI) | aHR (95% CI) | xsp | ||
| ED diagnosis | 1,298,331 | 2,972 (0.23) | Ref | Ref | 0.090 | 3,262 (0.25) | Ref | Ref | 0.051 | 1,098 (0.08) | Ref | Ref | 0.140 | 7,046 (0.55) | Ref | Ref | 0.002 |
| ED + PDE5i | 615,838 | 2,079 (0.34) | 1.07 (1.01 – 1.55) | 1.02 (0.96 – 1.07) | 2,523 (0.41) | 1.10 (1.05 – 1.16) | 1.04 (0.99 – 1.10) | 751 (0.12) | 1.03 (0.94 – 1.13) | 0.97 (0.88 – 1.06) | 5,134 (0.84) | 1.08 (1.04 – 1.12) | 1.001 (0.97 – 1.04) | ||||
| ED + ICI/urethral supp | 14,780 | 84 (0.57) | 1.57 (1.26 – 1.95) | 1.32 (1.06 – 1.64) | 120 (0.81) | 1.84 (1.53 – 2.20) | 1.16 (0.97 – 1.39) | 41 (0.28) | 2.02 (1.48 – 2.76) | 1.32 (0.97 – 1.81) | 240 (1.6) | 1.80 (1.59 – 2.05) | 1.25 (1.10 – 1.42) | ||||
| ED + PP | 9,313 | 43 (0.46) | 1.15 (0.85 – 1.55) | 0.97 (0.72 – 1.31) | 103 (1.1) | 2.19 (1.80 – 2.67) | 1.24 (1.01 – 1.50) | 30 (0.32) | 2.10 (1.46 – 3.02) | 1.25 (0.87 – 1.80) | 169 (1.8) | 1.79 (1.54 – 2.09) | 1.16 (0.997 – 1.35) | ||||
| BPH diagnosis | 1,178,235 | 3,095 (0.26) | Ref | Ref | 0.110 | 3,744 (0.32) | Ref | Ref | <0.001 | 1,301 (0.11) | Ref | Ref | 0.012 | 7,804 (0.67) | Ref | Ref | <0.001 |
| BPH + PDE5i | 175,725 | 720 (0.41) | 1.07 (0.99 – 1.16) | 1.09 (0.999 – 1.18) | 994 (0.57) | 1.13 (1.06 – 1.22) | 1.14 (1.06 – 1.23) | 300 (0.17) | 1.06 (0.94 – 1.21) | 1.07 (0.94 – 1.21) | 1,914 (1.1) | 1.09 (1.04 – 1.15) | 1.09 (1.04 – 1.15) | ||||
| BPH + 5a reductase/alpha blocker | 744,755 | 2,820 (0.38) | 1.13 (1.07 – 1.18) | 1.05 (0.998 – 1.11) | 5,198 (0.70) | 1.63 (1.56 – 1.70) | 1.11 (1.06 – 1.16) | 1,461 (0.20) | 1.39 (1.29 – 1.50) | 1.01 (0.93 – 1.09) | 9,055 (1.2) | 1.40 (1.36 – 1.44) | 1.08 (1.04 – 1.11) | ||||
| BPH + Surgery | 76,724 | 378 (0.49) | 1.19 (1.07 – 1.32) | 1.07 (0.96 – 1.19) | 782 (1.0) | 1.86 (1.72 – 2.00) | 1.10 (1.02 – 1.19) | 260 (0.34) | 1.95 (1.70 – 2.22) | 1.24 (1.08 – 1.42) | 1,365 (1.8) | 1.63 (1.54 – 1.73) | 1.12 (1.06 – 1.19) | ||||
Figure 1.
Forrest plot representing hazard ratios (HR) and 95% confidence intervals for incident ocular events among erectile dysfunction (ED) patients stratified by ED treatment.
Among patients diagnosed with BPH, 3,095 (0.26%), 3,744 (0.32%), and 1,301 (0.11%) experienced SRD, RVO, and ION, respectively. Among patients diagnosed with BPH and treated with PDE5i, 720 (0.41%), 994 (0.57%), and 300 (0.17%) experienced SRD, RVO, and ION, respectively. Among patients diagnosed with BPH and treated with 5a reductase/alpha blocker, 2,820 (0.38%), 5,198 (0.70%), and 1,461 (0.20%) experienced SRD, RVO, and ION, respectively. Among patients diagnosed with BPH and treated surgically, 378 (0.49%), 782 (1.0%), and 260 (0.34%) experienced SRD, RVO, and ION, respectively. On adjusted Cox regression analysis, PDE5i use was not associated with SRD and ION when compared to BPH diagnosisis and other BPH treatments. In contrast, PDE5i use was associated with RVO when compared to BPH diagnosisis and other BPH treatments (HR: 1.14; 95% CI: 1.06 −1.23). Moreover, BPH patients using 5a reductase/alpha blocker (HR: 1.11; 95% CI: 1.06–1.16) or surgically treated (HR: 1.10; 95% CI: 1.02–1.19) had a higher risk of RVO when compared to untreated BPH patients and those treated with other modalities. BPH patients treated with PDE5i (HR: 1.09; 95% CI: 1.04–1.15), 5a reductase/alpha blocker (HR: 1.08; 95% CI: 1.04–1.11), and surgery (HR: 1.12; 95% CI: 1.06–1.19) had a higher risk of developing any ocular events when compared to men diagnosed with untreated BPH patients. Figure 2 depicts the HRs for all BPH patients populations predicting any ocular event.
Figure 2.
Forrest plot representing hazard ratios (HR) and 95% confidence intervals for incident ocular events among benign prostatic hyperplasia (BPH) patients stratified by BPH treatment.
When patients were stratified according to age (Sup. Table 1), number of comorbidities (Sup. Table 2), and number of prescriptions (Sup. Table 3) the results were not meaningfully altered.
4. Discussion
The current study did not identify an association between PDE5i use and the occurrence of SRD, RVO, or ION. When compared to other ED or BPH treatments, patients using PDE5i did not have an increased risk of ocular events. Moreover, as treatment intensity increased for ED or BPH, the risk of ocular events increased implying a shared etiology rather than a treatment related side effect.
Among the different side effects of PDE5i, SRD, RVO, or ION are among the least investigated. Most of the evidence on the three ocular outcomes linked to PDE5i usage has been published in case reports or case series, which cannot demonstrate a causal link17. Nathoo et al investigated the role of PDE5i in the development of nonarteritic anterior ischemic optic neuropathy (NAION) in males. In their database, 1,109 cases of NAION were discovered and matched to 1,237,290 controls. Neither the adjusted rate ratio for PDE5i usage overall in the year before NAION nor the PDE5i use recently in the 30 days prior to the NAION showed any significant association. Similarly, individual PDE-5 inhibitor results failed to reach statistical significance18. Similarly, a review of over 14,000 sildenafil cases found no evidence of a higher incidence of NAION than in the general population19. In a 12-week, double-blind, randomized, placebo-controlled, Phase-III trial, Wirostko et al. investigated the ocular safety of chronic sildenafil use. 277 men with idiopathic pulmonary arterial hypertension were given oral sildenafil (20, 40, or 80 mg) or a placebo, and they were monitored for 18 months. Ophthalmic examinations, visual function tests, patient reports of adverse events, and adverse events reported by investigators were all used to examine ocular adverse events. None of the sildenafil-taking groups had NAION at the end of the 18-month period20. However, a recent paper by Etminan et al used data from 213 033 men receiving PDE5is. The group analyzed the risk of SRD, RVO, and ION and reported an adjusted incidence rate ratio for the composite endpoints of any one of the three outcomes of 1.85 (95% CI, 1.41–2.42) 11.
Importantly, the most common indication of PDE5i (i.e. ED) shares risk factors with the ocular outcomes reported. Just as vasculogenic disease remains the most common cause of erectile dysfunction, vascular and metabolic disease are well-known risk factors for SRD, RVO, and ION2,13,21,22. For this reason, it is challenging to determine whether the eiology of any reported outcome is the disease (or risk factor) or the treatment, especially when only one treatment for a disease is examined. In order to address this potential bias, we investigated not only different treatments for ED but a different indications as well. BPH risk factors and pathophysiology are extremely different from ED’s ones and permitted us to test the association between PDE5i in a different context. When examined through this lense, the current report did not find an association between PDE5i use and SRD, RVO, or ION. Moreover, the treatment of ED generally progresses from least invasive (e.g. lifestyle changes, oral medications) to more invasive (e.g. urethral suppositories, penile injections, surgical penile prosthesis) which mirrors effectiveness of therapies for more severe disease. In general, all treatments followed a similar pattern to PDE5i use.
Our study is not devoid of limitations. The main drawback was that we could only confirm actual intake based on data regarding drug dispensation. Additionally, because we did not have access to the patient’s medical records, the outcome definitions were diagnostic/treatment appropriate granularity in some cases. Next, using an insurance claims database composed of individuals with employed based insurance, may limit generalizability. Moreover, due to the small number of events, we were unable to separately assess the risk for each PDE5I’s influence on the study outcomes. In addition, there is a chance of measurement error for the covariates adjusted for, as with all pharmacoepidemiologic studies that use health claims data.
5. Conclusions
In our study, we did not observe any consistent associations between PDE5i use and any ocular adverse events (SRD, RVO, and ION). Moreover, patients using PDE5i did not have an increased risk of ocular events even when compared to other ED treatments and to patients diagnosed with BPH and the use of various BPH treatments.
Supplementary Material
References
- 1.Feldman HA, Goldstein I, Hatzichristou DG, Krane RJ, McKinlay JB. Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J Urol. 1994;151(1):54–61. doi: 10.1016/s0022-5347(17)34871-1 [DOI] [PubMed] [Google Scholar]
- 2.Yafi FA, Jenkins L, Albersen M, et al. Erectile dysfunction. Nat Rev Dis Primer. 2016;2:16003. doi: 10.1038/nrdp.2016.3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Jackson G, Montorsi P, Adams MA, et al. Cardiovascular aspects of sexual medicine. J Sex Med. 2010;7(4 Pt 2):1608–1626. doi: 10.1111/j.1743-6109.2010.01779.x [DOI] [PubMed] [Google Scholar]
- 4.Salonia A, Capogrosso P, Clementi MC, Castagna G, Damiano R, Montorsi F. Is erectile dysfunction a reliable indicator of general health status in men? Arab J Urol. 2013;11(3):203–211. doi: 10.1016/j.aju.2013.07.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Besiroglu H, Otunctemur A, Ozbek E. The relationship between metabolic syndrome, its components, and erectile dysfunction: a systematic review and a meta-analysis of observational studies. J Sex Med. 2015;12(6):1309–1318. doi: 10.1111/jsm.12885 [DOI] [PubMed] [Google Scholar]
- 6.Pizzol D, Smith L, Fontana L, et al. Associations between body mass index, waist circumference and erectile dysfunction: a systematic review and META-analysis. Rev Endocr Metab Disord. 2020;21(4):657–666. doi: 10.1007/s11154-020-09541-0 [DOI] [PubMed] [Google Scholar]
- 7.Isidori AM, Buvat J, Corona G, et al. A critical analysis of the role of testosterone in erectile function: from pathophysiology to treatment-a systematic review. Eur Urol. 2014;65(1):99–112. doi: 10.1016/j.eururo.2013.08.048 [DOI] [PubMed] [Google Scholar]
- 8.Corona G, Petrone L, Mannucci E, et al. Assessment of the relational factor in male patients consulting for sexual dysfunction: the concept of couple sexual dysfunction. J Androl. 2006;27(6):795–801. doi: 10.2164/jandrol.106.000638 [DOI] [PubMed] [Google Scholar]
- 9.Salonia A, Bettocchi C, Boeri L, et al. European Association of Urology Guidelines on Sexual and Reproductive Health-2021 Update: Male Sexual Dysfunction. Eur Urol. 2021;80(3):333–357. doi: 10.1016/j.eururo.2021.06.007 [DOI] [PubMed] [Google Scholar]
- 10.Hernandez I, Gul Z, Gellad WF, Davies BJ. Marked Increase in Sales of Erectile Dysfunction Medication During COVID-19. J Gen Intern Med. 2021;36(9):2912–2914. doi: 10.1007/s11606-021-06968-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Etminan M, Sodhi M, Mikelberg FS, Maberley D. Risk of Ocular Adverse Events Associated With Use of Phosphodiesterase 5 Inhibitors in Men in the US. JAMA Ophthalmol. Published online April 7, 2022. doi: 10.1001/jamaophthalmol.2022.0663 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Li AS, Pomeranz HD. Food and Drug Administration Adverse Event Reports of Retinal Vascular Occlusions Associated With Phosphodiesterase Type 5 Inhibitor Use. J Neuro-Ophthalmol Off J North Am Neuro-Ophthalmol Soc. 2016;36(4):480–481. doi: 10.1097/WNO.0000000000000450 [DOI] [PubMed] [Google Scholar]
- 13.Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye Res. 2009;28(1):34–62. doi: 10.1016/j.preteyeres.2008.11.002 [DOI] [PubMed] [Google Scholar]
- 14.Belladelli F, Del Giudice F, Eisenberg ML. In Defense of Phosphodiesterase 5 Inhibitors. JAMA Ophthalmol. Published online July 21, 2022. doi: 10.1001/jamaophthalmol.2022.2612 [DOI] [PubMed] [Google Scholar]
- 15.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–383. doi: 10.1016/0021-9681(87)90171-8 [DOI] [PubMed] [Google Scholar]
- 16.Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130–1139. doi: 10.1097/01.mlr.0000182534.19832.83 [DOI] [PubMed] [Google Scholar]
- 17.Margo CE, French DD. Ischemic optic neuropathy in male veterans prescribed phosphodiesterase-5 inhibitors. Am J Ophthalmol. 2007;143(3):538–539. doi: 10.1016/j.ajo.2006.10.006 [DOI] [PubMed] [Google Scholar]
- 18.Nathoo NA, Etminan M, Mikelberg FS. Association between phosphodiesterase-5 inhibitors and nonarteritic anterior ischemic optic neuropathy. J Neuro-Ophthalmol Off J North Am Neuro-Ophthalmol Soc. 2015;35(1):12–15. doi: 10.1097/WNO.0000000000000186 [DOI] [PubMed] [Google Scholar]
- 19.Bertero E, Montorsi F. Safety of sildenafil citrate: review of 67 double-blind placebo-controlled trials and the postmarketing safety database: Giuliano F, Jackson G, Montorsi F, Martin-morales A, and Raillard P. J Sex Med. 2014;11(4):885–887. doi: 10.1111/jsm.12513 [DOI] [PubMed] [Google Scholar]
- 20.Wirostko BM, Tressler C, Hwang LJ, Burgess G, Laties AM. Ocular safety of sildenafil citrate when administered chronically for pulmonary arterial hypertension: results from phase III, randomised, double masked, placebo controlled trial and open label extension. BMJ. 2012;344:e554. doi: 10.1136/bmj.e554 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Pozzi E, Capogrosso P, Boeri L, et al. Longitudinal Risk of Developing Cardiovascular Diseases in Patients With Erectile Dysfunction-Which Patients Deserve More Attention? J Sex Med. 2020;17(8):1489–1494. doi: 10.1016/j.jsxm.2020.03.012 [DOI] [PubMed] [Google Scholar]
- 22.Muraoka Y, Murakami T, Nishijima K, et al. Association between retinal venular dilation and serous retinal detachment in diabetic macular edema. Retina Phila Pa. 2014;34(4):725–731. doi: 10.1097/IAE.0b013e3182a48732 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.