Embed Size (px)
FEATURED NEW INVESTIGATORTestosterone, thrombophilia, thrombosis
JOEL FREEDMAN, CHARLES J. GLUECK, MARLOE PRINCE, RASHID RIAZ, and PING WANG
Joel Freedman, MD, is a third yeat the Jewish Hospital Internal Marticle is based on a presentationMeeting of the Central SocietyResearch and the Midwestern SeMedical Research, held in ChicagFrom the Jewish Hospital Interna
Cincinnati, Ohio; Cholesterol, Met
Jewish Hospital of Cincinnati, Cinc
We screened previously undiagnosed thrombophilia (V Leiden–prothrombin muta-tions, Factors VIII and XI, homocysteine, and antiphospholipid antibody [APL]syndrome) in 15 men and 2 women with venous thromboembolism (VTE) orosteonecrosis 7 months (median) after starting testosterone therapy (TT), gel(30–50 mg/d), intramuscular (100–400 mg/wk), or human chorionic gonadotropin(HCG) (6000 IU/wk). Thrombophilia was studied in 2 healthy control groups withoutthrombosis (97 normal controls, 31 subjects on TT) and in a third control group(n 5 22) with VTE, not on TT. Of the 17 cases, 76% had $1 thrombophilia vs 19% of97 normal controls (P , 0.0001), vs 29% of 31 TT controls (P 5 0.002). Cases differedfrom normal controls by Factor V Leiden (12% vs 0%, P 5 0.021), by high Factor VIII(.150%) (24% vs 7%, P 5 0.058), by high homocysteine (29% vs 5%, P 5 0.007),and from both normal and TT controls for APL syndrome (18% vs 2%, P 5 0.023, vs0%, P 5 0.04). Despite adequate anticoagulation with TT continued after the firstdeep venous thrombosis–pulmonary embolus (DVT-PE), 1 man sustained 3 DVT-PEs5, 8, and 11 months later and a second man had 2 DVT-PEs 1 and 2 months later.Of the 10 cases with serum T measured on TT, 6 (60%) had supranormal T (.800 ng/dL) and of 9 with estradiol measured on TT, 7 (78%) had supranormal levels(.42.6 pg/mL). TT interacts with thrombophilia leading to thrombosis. TT continuationin thrombophilic men is contraindicated because of recurrent thrombi despite anti-coagulation. Screening for thrombophilia before starting TT should identify subjects athigh risk for VTE with an adverse the risk to benefit ratio for TT. (Translational Research2015;165:537–548)
Abbreviations:APL¼antiphospholipid antibody; DVT¼deep venous thrombosis; E2¼ estradiol;FV ¼ Factor V; HRT ¼ hormone replacement therapy; ON ¼ osteonecrosis; PE ¼ pulmonaryembolus; PTG ¼ prothrombin gene mutation; T ¼ testosterone; TT ¼ testosterone therapy; VTE¼ venous thromboembolism
A ndrogen use in men aged $40 has increasedmore than 3-fold from 0.81% in 2001 to2.91% in 2011.1 The broad use of testosterone
therapy (TT) may have major public health ramifica-
ar resident in Internal Medicineedicine Residency Program. Hisgiven at the Combined Annualfor Clinical and Translationalction American Federation foro, Ill, on April 2014.l Medicine Residency Program,
abolism, and Thrombosis Center,
tions, given recent reports of thrombotic2-7 andcardiovascular disease (CVD) events8-10 associatedwith TT. Despite lessons learned about venousthromboembolism (VTE) and CVD associated withsex-hormone therapy in postmenopausal women from
Submitted for publication September 30, 2014; revision submitted
December 16, 2014; accepted for publication December 17, 2014.
Reprint requests: Charles J. Glueck, Cholesterol, Metabolism, and
Thrombosis Center, Suite 430, 2135 Dana Avenue, Cincinnati, OH
45207; e-mail: [email protected].
1931-5244/$ - see front matter
� 2015 Elsevier Inc. All rights reserved.
AT A GLANCE COMMENTARY
Freedman J, et al.
When testosterone therapy (TT) is given to men
and women, thrombotic events occur, reflecting
thrombophilic interactions between TT and famil-
ial and acquired thrombophilia-hypofibrinolysis.
TT appears to interact with thrombophilia leading
to thrombosis. TT continuation in thrombophilic
men is contraindicated because of recurrent
thrombi despite adequate anticoagulation.
Screening for thrombophilia (PCR measures of
the Factor V Leiden, and prothrombin gene muta-
tions, Factors VIII and XI, homocysteine, and
antiphospholipid antibodies) before starting TT
should identify subjects at high risk for venous
thromboembolism with an adverse risk/benefit ra-
tio for TT.
Translational Research538 Freedman et al May 2015
the Women’s Health Initiative,11,12 TT is oftenindiscriminately prescribed to middle-age, obese menwithout understanding of its long-term risks.13 Manyof these men do not meet highly specific criteria fordiagnosis and therapy of hypogonadism.14 Testosterone(T) levels fall with increasing age,15,16 with chronicdisease17,18 and obesity,17 and rise with smoking.19,20
Part of the problem of increased TT use lies indetermination of an age-specific lower normal rangefor T, because most normal ranges come from healthyyounger men.21 There are also differences in T assaymethods22,23 and recognition that adverse musclesymptoms occur at different T levels in differentsubjects.24
Previously, in aggregate,2-7 we have described VTEdeveloping after 5 months (median) in 42 patients onTT, 38 men and 4 women, including 27 with deepvenous thrombosis–pulmonary embolism (DVT-PE),12 with osteonecrosis (ON),4,25,26 1 with centralretinal vein occlusion (CRVO), 1 with amaurosisfugax, and 1 with spinal cord infarction. None of the42 cases in our previous studies2-7 had increasedhemoglobin or uncontrolled hypertension during TT,which might have contributed to their thrombi. Ofthe 40 cases having studies of thrombophilia, 39 werefound to have previously undiagnosed thrombophilia-hypofibrinolysis, including 28% heterozygous for theFactor V (FV) Leiden mutation, 28% with high Factor
VIII, and 15% with high Factor XI.7 In 8 men whoseTT was continued, second thrombotic event occurreddespite adequate anticoagulation with warfarin.7
Beyond interacting with familial and acquired throm-bophilia,2-7 TT is associated with physiological changesthat predispose to clotting and thrombosis includinghypertension,27 increased hemoglobin,28 low high-density lipoprotein cholesterol,27,28 hyperviscosity,and platelet aggregation.28-30 Dihydrotestosteroneenhances monocyte activation,31 further promotingacute coronary events.32 TT also increases circulatingestrogens33 that subsequently play a role in thromboticevents.7 Oral contraceptives and hormone replacementtherapy (HRT)34,35 have been identified as risk factorsfor VTE in women. Given that T is converted byaromatization to estradiol (E2),36 it may be prothrom-botic by the same mechanism as reported in women,where HRT interacts with the FV Leiden mutation to in-crease risk of VTE.37
In June 2014, on the basis of postmarketing surveil-lance reports including our studies,2-7 both the USFood and Drug Administration (FDA)38 and CanadaHealth39 added a warning regarding the risks of VTEto the label of all T products. VTE, particularly PE, isassociated with significant mortality risk.40 The FDAhad previously warned about a TT-associated increasein VTE in men with a diagnosis of increased hemoglo-bin.28 However, the June 2014 FDA warning38 wasbased on reports of VTE in men without increased he-moglobin.In the present report, our specific aim was to further
focus on thrombotic events in 17 newly reported pa-tients on TT (1 receiving HCG injections as a form ofTT41), subsequently found to have previously undiag-nosed familial and acquired thrombophilia. Our secondspecific aim was to assess thrombophilia in our totalcohort of 57 patients who sustained VTE on TT, 17newly reported and 40 previously reported.7 Our thirdspecific aim was to compare thrombophilia in 57 caseswho developed VTE on TT, in 97 healthy normal con-trols, in 31 men receiving TT without VTE, and in 22cases with VTE, not taking TT.
MATERIALS AND METHODS
Patients. The procedures followed were in accordancewith the ethical standards of the institutional reviewboard of the Jewish Hospital, Cincinnati, Ohio, whichapproved the research protocol. The protocol was car-ried out with the understanding, and signed informedconsent was taken from each participant.We excluded patients and controls whose VTE was
associated with cancer, polycythemia vera, recent soft
Fig 1. Percent of subjects with no thrombophilia or $1 abnormality
of thrombophilia (V Leiden mutation, prothrombin gene mutation,
high Factors VIII, XI, homocysteine, and antiphospholipid antibody
[APL] syndrome), in 17 cases with venous thromboembolism on
testosterone therapy (TT-VTE), 97 healthy normal controls, 31 con-
trols on TT without VTE (TT controls), and 22 controls with VTE
not on TT (NTT-VTE). Comparisons were made by Fisher’s exact
test (significance level, P , 0.05).
Fig 2. Differences in thrombophilias between 17 venous thromboem-
bolism on testosterone therapy (TT-VTE) cases and 97 normal
controls, 31 TT controls, and 22 controls with VTE not on TT
(NTT-VTE). Comparisons were made by Fisher’s exact test (signifi-
cance level, P , 0.05).
Translational ResearchVolume 165, Number 5 Freedman et al 539
tissue trauma, bone fracture, hip-knee-foot surgery,airline flights .8 hours, or extensive immobilization.Fifteen men and 2 women were studied in their serial
order of referral in the last 14 months in our ThrombosisCenter by their family physicians-urologists because ofthrombotic events. TT had been prescribed by the refer-ring family physicians-urologists, and, if not alreadystopped, it was discontinued by us before our assessmentof thrombophilia and hypofibrinolysis. Measurement ofthrombophilia-hypofibrinolysis was carried out at least2 months after the initial thrombotic event (Figs 1and 2). We focused on 6 thrombophilias previouslyimplicated7 in a TT-thrombophilia interaction for VTE(FV Leiden25 and prothrombin gene mutations[PTGs], high Factors VIII and XI, high homocysteine,and the antiphospholipid antibody [APL] syndrome).The 17 newly acquired cases of the present study were
added to 40 previously reported cases7 to provide 57cases for comparison with controls (Figs 3 and 4).At study entry, a detailed history42-44 for VTE, DVT,
PE, recurrent miscarriage, pre-eclampsia, eclampsia,and hemolysis-elevated liver enzymes-low plateletcount syndrome was taken for all the cases and theirfirst-degree relatives, along with a history of cigarettesmoking.In the 17 cases and 31 controls receiving TT but
without VTE, total serum T and E2 were measured inblood drawn in the morning.
Controls. Six tests for thrombophilia (FV Leiden,PTG, Factors VIII and XI, homocysteine, and APL syn-drome) were carried out in 110 healthy controls (46 men
and 64 women), free of VTE, not on TT, and in a secondcontrol group, 56 healthy subjects (52 men and 4women) who took T (TT controls), with neither controlgroup having DVT-PE or ON. A third control groupincluded 22 cases referred to us for evaluation ofVTE, not taking TT (NTT-VTE controls).In 31 TT controls and the 22 NTT-VTE controls, at
study entry, a detailed history42-44 for VTE, DVT, PE,recurrent miscarriage, pre-eclampsia, eclampsia, andhemolysis-elevated liver enzymes-low platelet countsyndrome was taken for all probands and their first-degree relatives, along with a history of cigarette smok-ing.In our analysis to optimize comparability of cases and
controls, we included cohort of only those cases andcontrols having complete data for the 6 tests of throm-bophilia (FV, PTG, Factors VIII and XI, homocysteine,and APL syndrome), including 97 of the 110 normalcontrols, 31 of the 56 TT controls, and all 22 of theVTE controls (Table I).
Laboratory assessment of thrombophilia. Polymerasechain reaction assays. Polymerase chain reaction(PCR) measures of thrombophilia (G1691A FV Leidenand G20210A prothrombin) were performed in all casesand controls using previously published methods25,26 bylaboratory staff blinded to the participants’ status(diagnosis and severity of disease).
Serologic measures of thrombophilia. Serologic mea-sures of thrombophilia included Factors VIII and XI,homocysteine, and the APL syndrome (anticardiolipinantibodies [IgG, IgM], the lupus anticoagulant, andanti-beta2-glycoprotein). Previously published estab-lished methods were used.25 High homocysteine was
Fig 3. Percent of subjects with no thrombophilia or$1 thrombophilia
(V Leiden mutation, prothrombin gene mutation, high Factors VIII,
XI, homocysteine, and antiphospholipid antibody [APL] syndrome)
in 57 cases with venous thromboembolism on testosterone therapy
(TT-VTE), 97 normal controls, 31 controls on TT without VTE (TT
controls), and 22 controls with VTE not on TT (NTT-VTE). Compar-
isons were made by Fisher’s exact test (significance level, P , 0.05).
Fig 4. Difference in thrombophilias between 57 venous thromboem-
bolism on testosterone therapy (TT-VTE) cases and 97 normal
controls, 31 TT controls, and 22 controls with VTE not on TT
(NTT-VTE). Comparisons were made by Fisher’s exact test (signifi-
cance level, P , 0.05).
Translational Research540 Freedman et al May 2015
identified by levels .the 95th percentile45 for controls.The 3 components of the APL syndromewere measuredby enzyme-linked immunosorbent assay using previ-ously published methods.46-48
Measures of T and E2. Total T, E2, and free E2 weremeasured by LabCorp in a single laboratory. Total Tand E2 were measured by electrochemiluminescenceimmunoassay.49
Statistical methods. All statistical analyses were doneusing Statistical Analysis System version 9.4. Wilcoxonnonparametric tests were used for comparisons ofcontinuous variables. For categorical variables, Fisher’sexact test was used. Analysis of variance was used aftercovariance adjusting for age to compare homocysteinelevels in patients vs normal controls.
Patients and controls. By selection, none of the newlyreported 17 cases, the 40 previously reported cases,7 orthe 22 patients with NTT-VTE control group hadpathoetiologic causes for DVT-PE (cancer, increasedhemoglobin, antecedent trauma or surgery, recentairline travel longer than 8 hours, or extendedimmobilization).Because we did not initiate TT in either the TT-treated
patients with VTE or the TT controls without VTE, wedo not know the etiology of the hypogonadism in the 2TT-treated groups.Fifteenmen and2womenwere allCaucasianwith ages
ranging from 23 to 80, mean 546 16, median 55 years,
12 of whom were aged #60 (Table II). None of the pa-tients had sustained thrombotic events before TT.The 17 cases were older than the 97 normal controls
(43 6 12, median 42, P 5 0.006), but their age didnot differ from the 31 TT controls (52 6 12, median52) or from the 22 NTT-VTE controls (596 17, median63), P . 0.05 for both.Of the 17 cases 6 (35%) had normal body mass index
(BMI) (18.5 to ,25 kg/m2), 6 (35%) were overweight(25 to ,30), and 5 (30%) were obese (BMI $ 30)(Table II). Of the 97 normal controls, 40% had normalBMI, 28% were overweight, and 32% were obese. Ofthe 31 TT controls, 13% had normal BMI, 48% wereoverweight, and 39% were obese. Of the 22 VTE con-trols, 10% had normal BMI, 14% were overweight,and 76% were obese. BMI in the 17 cases did not differfrom normal or TT controls (P . 0.2 for both), but the22 VTE controls were heavier than the 17 cases (median33.1 vs 27.1 kg/m2, P 5 0.01, Wilcoxon).Of the 17 cases, 10 had measures of serum Twhen on
TT, and serum T was supranormal ($800 ng/dL) in 6(60%) (Table II), compared with 8 of 29 (28%) TT con-trols (P5 0.12). Median T in the cases was higher thanin the TT controls (1198 vs 460 ng/dL, P 5 0.018). Ofthe 17 cases, 9 had measures of E2 when on TT, and ofthese 9, 7 (78%) had high serum E2 $42.6 pg/mL(Table II), vs 11 of 28 (39%) TT controls (P 5 0.06).Median E2 in cases differed from TT controls (52.5 vs35.1 pg/mL, P 5 0.053).Family history of VTE was positive for 1 of the 17
cases (6%), not different (P. 0.05) from the 31 TT con-trols (0%) or the 22 VTE controls (10%).
Translational ResearchVolume 165, Number 5 Freedman et al 541
Of the 17 cases, 1 smoked (6%), not different (P. 0.05)from 0 of 31 TT controls (0%) or 3 of 22 NTT-VTE con-trols (14%).
Thrombotic events in cases. Of the 15 men and 2women, 6 (case IDs 1, 3, 5, 10, 11, and 12, Table II)had DVT alone, 4 (case IDs 2, 6, 8, and 9) had DVT-PE, 1 (case ID 4) had PE alone, 1 (case ID 7) hadCRVO and DVT, 2 (case ID14 and 15) had CRVOalone, and 3 (case ID 13, 16, and 17) had ON, In 15men and 2 women, thrombotic events occurred7 months (median) after starting TT, with 9 eventsfrom 1 to 7 months after starting TT and 8 eventsfrom 12 to 36 months (Table II).After the initial thrombotic event, TTwas continued in
2 men (case IDs 2 and 9), with concurrent anticoagula-tion (Table II). In case ID 2, despite adequate anticoagu-lation with warfarin (international normalizedratio $ 3.0), DVT-PE occurred a second, third, andfourth time 5, 8, and 11 months after the initial event(Table II). In case ID 9, after an initial right arm and sub-clavian thrombus and PE, second and third DVT-PEoccurred 1 and 2 months after the initial event, despitefull dose Xarelto (20 mg/d) (Table II).
Thrombophilia. Of the 17 cases, 13 (76%) had $1thrombophilia vs 19% in 97 normal controls(P , 0.0001), vs 29% in 31 TT controls (P 5 0.002),vs 55% in the 22 VTE controls (P 5 0.2) (Fig 1).Healthy normal and TT controls did not differ(P 5 0.2) in having $1 thrombophilia 19% vs 29%(Fig 1, Table I). VTE controls were more likely thanhealthy normal controls to have $1 thrombophilia55% vs 19% (P 5 0.002), but did not differ(P 5 0.09) from TT controls 29% (Fig 1, Table I).The 17 cases differed from normal controls by FV
Leiden (12% vs 0%, P 5 0.021), by high Factor VIII(.150%) (24% vs 7%, P 5 0.058) and by high homo-cysteine (29% vs 5% P 5 0.007) (Fig 2, Table III).After covariance adjusting for age, least squaresmean mean homocysteine in the 17 cases(11.5 6 0.9 mmol/L) was higher than in the normalcontrols (8.2 6 0.4), P 5 0.001. The 17 cases differedfrom both normal and TT controls for the APL syn-drome (18% vs 2% in normal controls, P 5 0.023; vs0% TT controls, P 5 0.039) (Fig 2, Table III). The17 cases did not differ (P . 0.05) from the 22 VTEcontrols for Factor VIII (24% vs 50%), for high homo-cysteine (29% vs 23%), or for the APL syndrome (18%vs 5%) (Fig 2, Table III).Adding coagulation measures from 40 recently pub-
lished cases7 to the current 17 cases allowed comparisonof 57 patients (53 men, 4 women with VTE on TT) with97 normal controls, 31 TT controls, and 22 VTE con-trols. As displayed in Fig 3 and Table IV, 57 caseswere much more likely to have $1 of the 6 major
Table II. Deep venous thrombosis (DVT)-pulmonary embolus (PE)-central retinal vein occlusion (CRVO)-osteonecrosis (ON), which developed in 17
patients receiving testosterone therapy (TT) (1 receiving HCG injections as a form of TT)
Case ID Gender AgeBodymass
Event no., time between TT startand development of
Serum testosteronenormal 280–800 ng/dL
Serum estradiolnormal #42.6 pg/mL
DVT PE On Pre-Rx On Rx Pre-Rx On Rx
1 M 46 27.3 230 IM/wk 12 mo no PE 258 — 5.1 50.92 M 42 30.0 50 mg gel 1, 1 mo 1, 1 mo 120 — — 50.0
50 mg gel, warfarin 2, 6 mo 2, 6 mo50 mg gel, warfarin 3, 9 mo 3, 9 mo50 mg gel, warfarin 4, 12 mo 4, 12 mo
3 M 46 31.1 50 mg gel 3 mo no PE 192 120 42.7 —4 M 23 20.1 50 mg gel no DVT 7 mo 165 — — —5 M 75 26.7 100 mg IM 13 mo no PE 3.5 — — —6 M 80 23.5 30 mg gel 15 mo 15 mo — 750 — —7 M 72 53.0 200 IM 2/wk DVT 1 CRVO 3 mo no PE — 1500 — 1448 M 62 32.8 50 mg gel 36 mo 36 mo — 342 — 739 M 46 27.9 200 mg IM 2/wk Right arm subclavian PE
1, 14 mo 1, 14 mo 150 2000 5.9 —100 mg IM 1/wk
(Xarelto)2, 15 mo 2, 15 mo
100 mg IM 1/wk(Xarelto)
3, 16 mo 3, 16 mo
10 M 68 21.8 200 mg IM/wk 1 mo No PE 329 1500 21 7111 M 60 38.1 200 mg IM/wk 12 mo No PE 228 1198 — 5312 M 41 27.1 50 mg gel 36 mo No PE — 1467 — 5313 M 60 22.1 50 mg IM/wk No DVT No PE ON 12 mo — 1197 — 33.2, free E2 25.2
(UNL 18 pg/mL)14 M 32 22.7 HCG (6000 IU/wk)
anastrozole 1 mg/dCRVO 4 mo No DVT, no PE — 777 — 28.4
15 M 50 25.0 50 mg gel CRVO 3 mo No DVT, no PE 286 — — —16 F 40 28.0 50 mg gel 1 mo, shoulder
and knee— — — —
17 F 58 27.8 50 mg gel 6 mo, knee — — — —
Abbreviations: IM, intramuscular; UNL, upper normal limit.The bold values represent supranormal serum T or E2.
Time between TT start to development of event 10.5 6 10.8 months, median 5 7, range [1, 36].
Table III. Coagulation disorders in 17 cases with deep venous thrombosis, pulmonary embolism, central retinal vein occlusion, osteonecrosis on
testosterone therapy (TT) compared with 97 normal controls, to 31 controls on TT without VTE, and to 22 controls with VTE, not receiving TT
Case ID, Age (y) Factor V PTG Factor VIII Factor XI Homocysteine* APL syndrome$1 of 6 thrombophilia
abnormalitiesNormal CC CC #150% #150% Umol/L Negative
1, 46 CC CC 131 94 12.9 1 Y2, 42 CC CC 234 108 9.1 2 Y3, 46 CC CC 161 133 16.5 2 Y4, 23 CC TC 110 126 8.0 2 Y5, 75 CC CC 137 72 18.6 2 Y6, 80 CC CC 115 125 19.0 2 Y7, 72 CC TC 186 145 14.1 2 Y8, 62 CC CC 76 129 17.3 2 Y9, 46 CC CC 110 105 9.3 1 Y10, 68 CC CC 238 105 8.0 2 Y11, 60 CC CC 129 125 9.2 2 No12, 41 CC CC 130 135 8.0 1 Y13, 60 CC CC 71 113 8.0 2 No14, 32 TC CC 27 135 10.0 2 Y15, 50 CC CC 55 108 11.3 2 No16, 40 TC CC 121 N 10.0 2 Y17, 58 CC CC 145 112 9.4 2 NoPercentage of 17 cases
with thrombophilia2/17 (12%) 2/17 (12%) 4/17 (24%) 0/17 (0%) 5/17 (29%) 3/17 (18%) 13/17 (76%)
Percentage of 97 normalcontrols withthrombophilia vscases
0/97 (0%)P 5 0.021
7/97 (7%)P 5 0.058
5/97 (5%)P 5 0.007
2/97 (2%)P 5 0.023
18/97 (19%)P , 0.0001
Percentage of 31 TTcontrols without VTEvs cases
0/31 (0%)P 5 0.04
9/31 (29%)P 5 0.002
Percentage of 22 VTEcontrols without TT vscases
Abbreviations: APL, antiphospholipid antibody;CC, wild-type normal;NS, nonsignificant; PTG, prothrombin genemutation; TC, heterozygotemutant; TT, homozygous mutant; VTE, venous throm-boembolism.Abnormal represent by bold.*Homocysteine dated normal range (,laboratory 95th percentile):,13.5 mmol/L (beforeMarch 21, 2005),,12.0 (March 21, 2005 toMarch 27, 2006),,10.4 (March 28, 2006 to April 14, 2008),,11.4
(April 15, 2008 to November 14, 2008), and ,15 (since November 15, 2008).
Translational Research544 Freedman et al May 2015
thrombophilias (75%) than either normal controls(19%, P , 0.0001) or TT controls (29%, P , 0.0001).As displayed in Fig 4 and Table IV, 57 cases were
much more likely than normal controls to have FV Lei-den heterozygosity, high Factor VIII, and high homo-cysteine, and were also more likely to have V Leidenheterozygosity than the TT controls. VTE controls didnot differ significantly from the 57 TT-VTE casesfor FV 14% vs 23%, high Factor VIII 50% vs 26%(P 5 0.06), and high homocysteine 23% vs 18%(Fig 4, Table IV).Fifty-seven cases with VTE on TT did not differ
(P. 0.05) from the 31 TT controls or the 22 VTE con-trols for a positive family history for VTE 4% vs 0% vs10%, and did not differ for smoking 11% vs 0% vs 14%(P . 0.05 for all).
Testosterone and estradiol. Of the 57 cases with VTEon TT, 14 had measures of E2 when taking TT (1 takingHCG). E2 or free E2 was high (.42.6, .18 pg/mL) inthe 5 cases receiving intramuscular T, and E2 high in 7of 8 on T gel (P5 1.0). Of the 57 cases with VTE on TT,16 had measures of serum T when taking TT, 10 on Tgel, and 5 on intramuscular T, and 1 on HCG. SerumT was high (.800 ng/dL) in 5 of 5 cases on intramus-cular T and 2 of 10 cases on T gel, Fisher’s P 5 0.007.
As shown in the present report of 15 men and 2women who sustained thrombotic events on TT and inour previous report,7 TT appears to interact with throm-bophilia leading to thrombosis. Of the 17 cases, 76%had $1 abnormality of the 6 measured thrombophilias,higher than in 97 healthy controls (19%, P , 0.0001)and higher than in 31 TT controls (29%, P 5 0.0024).In the present study, FV Leiden (12%), high FactorVIII (24%), high homocysteine (26%), and the APLsyndrome (18%) were more common in the 17 caseswho sustained VTE on TT than in healthy normal con-trols, and cases also were more likely to have the APLsyndrome than in TT controls free of thrombotic events(18% vs 0%).In the elderly, mild reduction of renal function is
accompanied by increased serum homocysteine,50 andof the 5 cases with serum homocysteine .normals’95th percentile, the second highest levels were in menaged 75 and 80 years, whereas the remaining 3 subjectswith hyperhomocysteinemia were 46, 62, and 72 yearsold. However, after adjusting for age, mean homocyste-ine in the 17 cases differed from the 97 normal controls(11.5 vs 8.2 mmol/L, P 5 0.001).Adding the 17 new cases of the present report to those
40 previously reported,7 the 57 cases were more likelyto have $1 thrombophilia than 97 normal controls
Translational ResearchVolume 165, Number 5 Freedman et al 545
(75% vs 19%), or 31 TT controls (75% vs 29%). FV Lei-den heterozygosity (23%), high Factor VIII (26%), andhigh homocysteine (18%) were more common in 57cases than in 97 healthy normal controls (0%, 7%, and5%). FV Leiden heterozygosity was also more commonin the 57 cases than in TT controls (23% vs 3%).We do not believe that the high percentage of our pa-
tients with high Factor VIII (26%) in our study wasdriven by TT, because the percentage of subjects withhigh Factor VIII in our TT controls (13%) did not differ(P5 0.46) from that in 97 normal controls not receivingTT (7%). TT has not been reported to raise FactorVIII,51 and in a study of risk of thrombosis associatedwith high Factor VIII,52 TT was not recognized as acause of high Factor VIII.Mild-moderate homocysteinemia was much more
common in our 57 cases than in the 97 normal controls(18% vs 5%, P5 0.022). Mild-moderate homocysteine-mia in the range of our patients is a recognized risk fac-tor for venous thrombosis45,53-55 and particularly forocular vascular thrombosis.56
The percentage of our 57 patients who developedVTE after starting TT with $1 thrombophilia (75%)was higher than in the 97 normal controls (19%) orthe 31 TT controls (29%). Moreover, the percentageof our 57 patients with familial and acquired thrombo-philia (23%VLeiden heterozygotes, 7% PTG heterozy-gotes, 26% high Factor VIII, 11% high Factor XI, 18%high homocysteine, and 9% APL syndrome) was com-parable with or higher than reported studies of patientswith VTE not taking TT.57-61 Turan et al61 observed FVLeiden heterozygosity in 19% of patients with PE. In astudy of 3005 patients with VTE, Gadelha et al57 re-ported that 345 (11.4%) had the FV Leiden mutationand 8.7% had the PTG. Ivanov et al58 reported thatthe V Leiden mutation was present in 23.5% of patientswith PE and the PTG in 5.9%. In 44 patients with DVT,Caprini et al59 found 23% with the V Leiden mutation.In 318 patients with DVT, Ben Salah et al60 reported that22.6% had hereditary thrombophilia and 19.1% had theAPL syndrome.We speculate that TT is likely to result in VTE among
men or women3 with underlying familial or acquiredthrombophilia. Similarly, in women on HRT, the pres-ence of the FV Leiden mutation or high Factor VIII in-creases the risk of DVT 17-fold compared with womenon HRT without these thrombophilias.62 Moreover, thetime course between starting TT and VTE, 5 months(median) in our first 42 patients7 and 7 months (median)in the current 17 patients, is comparable with VTE inwomen receiving HRT, where VTE rates are highestin the first year of treatment and are much more pro-nounced in women at higher risk for VTE by virtue ofthrombophilia.35
In our study, family history of VTE was low, being6% in the 17 cases who developed VTE on TT, 0% inthe 31 TT controls, and 10% in the 22 VTE controls.These findings were congruent with previousstudies,42-44 emphasizing that family history is a weakpredictor of VTE. Cigarette smoking (6%) also didnot differ between our 17 patients with VTE on TT vsour 31 TT controls (0%) or our 22 VTE controls (14%).In men, most E2 comes from aromatization of T.63We
speculate that when exogenous T is aromatized to E2and E2-associated thrombophilia is superimposed on fa-milial and acquired thrombophilia, as in our present andprevious7 reports, thrombosis occurs. We have previ-ously reported that 45% of men receiving conventionalgel TT (50 mg/d) have high serum E2 ($42.6 pg/mL).4
In the present study, of the 9 men having measures of E2on TT, 7 (78%) had high E2. The high E2 may alsoreflect overtreatment with TT. Of the 10 men havingmeasures of T during TT, 6 (60%) had supranormalvalues (.800 ng/dL), and 5 of these 6 men also hadhigh E2 or free E2. We speculate that in the 17 caseswith VTE in the present study, high T and E2 on TTcontributed to their VTE, because median T in the 17cases was higher than in the 31 TT controls (1198 vs460 ng/dL, P 5 0.018). Median E2 in the 17 caseswas higher than in the 31 TT controls (52.5 vs35.1 pg/mL, P 5 0.053).After a thrombotic event, if TT is continued, concom-
itant and adequate anticoagulation does not appear toprevent recurrent thrombotic events.7 In the presentstudy, when TTwas continued along with optimal anti-coagulation with warfarin (international normalizedratio $ 3) or (Rivaroxaban 20 mg/d) after an initialthrombotic event, 1 patient sustained 3 subsequentDVT-PEs 5, 8, and 11 months after his first DVT-PEand a second patient had 2 DVT-PEs 1 and 2 months af-ter his first DVT-PE.There are similar time intervals between starting TT
and thrombotic events (median 5 months7) or cardiovas-cular events (�3 months8-10), and we speculate that theshort duration between starting TT and thrombotic andcardiovascular events may indicate a sharedthrombotic pathophysiology, because CVD events�3 months after starting TT cannot reflect aconventional arterial atherosclerotic event.Before starting TT, to facilitate prevention of VTE,
we recommend PCR measures of the FV Leiden andPTGs, as well as measures of Factors VIII and XI, ho-mocysteine, and the APL syndrome. In a parallelfashion, the issue of pretreatment thrombophiliascreening before giving exogenous estrogens has alsobeen raised in women.62,64-66
Our findings may have important clinical implica-tions, because VTE risk is an important determinant
Translational Research546 Freedman et al May 2015
of the benefit to risk ratio of TT. In women, for example,PE accounts for about one third of the incidence ofpotentially fatal VTE events associated with HRT,67
and HRT increases the risk of VTE by 2- to 3-fold.35
We studied only morbid VTE in the present study, alimitation, because mortal VTE might be associatedwith a larger percentage of subjects with thrombophilia.Because we did not initiate TT in either the TT-treatedpatients with VTE or TT controls without VTE, we donot know the etiology of the hypogonadism in the 2TT-treated groups. We had only limited serum T andE2 data in the 57 patients when on TT, with T high in7 of the 16 cases and E2 high in 12 of 14, and not enoughdata to determine whether intramuscular T led to higherT and E2 levels than T gel.An optimal study, paralleling the Women’s Health
Initiative studies in women11,12 would be prospective,placebo-controlled, and double-blind, with thromboticand CVD end points, and evaluation of increased hemo-globin, hypertension, prostate cancer, and osteoporosis.Alternatively outcomes of TT therapy could be studiedretrospectively using administrative health data fromhealth insurance sources provided that laboratory re-sults for thrombophilia measures were consistentlycoded. This study might allow an examination of aninteraction term (TT 3 thrombophilia) revealing, spec-ulatively, that TT is more likely to result in VTE amongthe subgroup of men with underlying thrombophilia.This study would also allow assessment of whethermen with low (,280 ng/dL) or normal serum T levels(280–800 ng/dL) before TT and those who reach supra-physiological levels of T (.800 ng/dL) during TT andthose have high E2 on TT (.42.6 pg/mL) are atincreased risk for thrombotic and CVD events with orwithout underlying thrombophilias.
Conflicts of Interest: All authors have read the jour-nal’s policy on disclosure of potential conflicts of inter-est and have none to declare.This work was supported by the Lipoprotein Research
Fund of the Jewish Hospital of Cincinnati.All authors have read the journal’s authorship agree-
ment and the manuscript has been reviewed andapproved by all named authors.There were no external sources of editorial support
for preparation of the manuscript.
Supplementary data related to this article can befound at http://dx.doi.org/10.1016/j.trsl.2014.12.003.
1. Baillargeon J, Urban RJ, Ottenbacher KJ, Pierson KS,
Goodwin JS. Trends in androgen prescribing in the United States,
2001 to 2011. JAMA Intern Med 2013;173:1465–6.
2. Glueck CJ, Richardson-Royer C, Schultz R, et al. Testosterone
therapy, thrombophilia-hypofibrinolysis, and hospitalization for
deep venous thrombosis-pulmonary embolus: an exploratory,
hypothesis-generating study. Clin Appl Thromb Hemost 2014;
3. Glueck CJ, Bowe D, Valdez A, Wang P. Thrombosis in three post-
menopausal women receiving testosterone therapy for low libido.
Womens Health (Lond Engl) 2013;9:405–10.
4. Glueck CJ, Goldenberg N, Budhani S, et al. Thrombotic
events after starting exogenous testosterone in men with previ-
ously undiagnosed familial thrombophilia. Transl Res 2011;
5. Glueck CJ, Richardson-Royer C, Schultz R, et al. Testosterone,
thrombophilia, and thrombosis. Clin Appl Thromb Hemost
6. Pandit RS, Glueck CJ. Testosterone, anastrozole, factor V Leiden
heterozygosity and osteonecrosis of the jaws. Blood Coagul Fibri-
7. Glueck CJ, Wang P. Testosterone therapy, thrombosis, thrombo-
philia, cardiovascular events. Metabolism 2014;63:989–94.
8. Basaria S, Coviello AD, Travison TG, et al. Adverse events asso-
ciated with testosterone administration. N Engl J Med 2010;363:
9. Vigen R, O’Donnell CI, Baron AE, et al. Association of testos-
terone therapy with mortality, myocardial infarction, and stroke
in men with low testosterone levels. JAMA 2013;310:1829–36.
10. Finkle WD, Greenland S, Ridgeway GK, et al. Increased risk of
non-fatal myocardial infarction following testosterone therapy
prescription in men. PLoS One 2014;9:e8595.
11. Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits
of estrogen plus progestin in healthy postmenopausal women:
principal results From the Women’s Health Initiative randomized
controlled trial. JAMA 2002;288:321–33.
12. Valdiviezo C, Lawson S, Ouyang P. An update on meno-
pausal hormone replacement therapy in women and cardio-
vascular disease. Curr Opin Endocrinol Diabetes Obes 2013;
13. Anawalt BD. Guidelines for testosterone therapy for men: how to
avoid a mad (t)ea party by getting personal. J Clin Endocrinol
14. Bhasin S, Cunningham GR, Hayes FJ, et al. Testosterone therapy
in men with androgen deficiency syndromes: an Endocrine Soci-
ety clinical practice guideline. J Clin Endocrinol Metab 2010;95:
15. Araujo AB, Esche GR, Kupelian V, et al. Prevalence of symptom-
atic androgen deficiency in men. J Clin Endocrinol Metab 2007;
16. Layton JB, Li D, Meier CR, et al. Testosterone lab testing and
initiation in the United Kingdom and the United States, 2000-
2011. J Clin Endocrinol Metab 2014;99:835–42.
17. Feldman HA, Longcope C, Derby CA, et al. Age trends in the
level of serum testosterone and other hormones in middle-aged
men: longitudinal results from the Massachusetts male aging
study. J Clin Endocrinol Metab 2002;87:589–98.
18. Wu FC, Tajar A, Pye SR, et al. Hypothalamic-pituitary-testicular
axis disruptions in older men are differentially linked to age and
modifiable risk factors: the European Male Aging Study. J Clin
Endocrinol Metab 2008;93:2737–45.
19. Svartberg J, Jorde R. Endogenous testosterone levels and smoking
in men. The fifth Tromso study. Int J Androl 2007;30:137–43.
Translational ResearchVolume 165, Number 5 Freedman et al 547
20. Wang W, Yang X, Liang J, et al. Cigarette smoking has a positive
and independent effect on testosterone levels. Hormones (Athens)
21. Lazarou S, Reyes-Vallejo L, Morgentaler A. Wide variability in
laboratory reference values for serum testosterone. J Sex Med
22. McShane LM, Dorgan JF, Greenhut S, Damato JJ. Reliability and
validity of serum sex hormone measurements. Cancer Epidemiol
Biomarkers Prev 1996;5:923–8.
23. YunYM,Botelho JC, Chandler DW, et al. Performance criteria for
testosterone measurements based on biological variation in adult
males: recommendations from the Partnership for the Accurate
Testing of Hormones. Clin Chem 2012;58:1703–10.
24. Finkelstein JS, Yu EW, Burnett-Bowie SA. Gonadal steroids and
body composition, strength, and sexual function in men. N Engl J
25. Glueck CJ, Freiberg RA, Wang P. Heritable thrombophilia-
hypofibrinolysis and osteonecrosis of the femoral head. Clin
Orthop Relat Res 2008;466:1034–40.
26. Glueck CJ, Freiberg RA, Boriel G, et al. The role of the factor V
Leiden mutation in osteonecrosis of the hip. Clin Appl Thromb
27. Spitzer M, Huang G, Basaria S, Travison TG, Bhasin S. Risks and
benefits of testosterone therapy in older men. Nat Rev Endocrinol
28. Fernandez-Balsells MM, Murad MH, Lane M, et al. Clinical re-
view 1: adverse effects of testosterone therapy in adult men: a sys-
tematic review and meta-analysis. J Clin Endocrinol Metab 2010;
29. Baskurt OK,Meiselman HJ. Iatrogenic hyperviscosity and throm-
bosis. Semin Thromb Hemost 2012;38:854–64.
30. Peerschke EI, Silver RT, Weksler BB, Yin W, Bernhardt B,
Varon D. Examination of platelet function in whole blood under
dynamic flow conditions with the cone and plate(let) analyzer: ef-
fect of erythrocytosis and thrombocytosis. Am J Clin Pathol 2007;
31. Death AK, McGrath KC, Sader MA, et al. Dihydrotestosterone
promotes vascular cell adhesion molecule-1 expression in male
human endothelial cells via a nuclear factor-kappaB-dependent
pathway. Endocrinology 2004;145:1889–97.
32. Pamukcu B, Lip GY, Devitt A, Griffiths H, Shantsila E. The role of
monocytes in atherosclerotic coronary artery disease. Ann Med
33. Tan RS, Cook KR, Reilly WG. High estrogen in men after inject-
able testosterone therapy: the low t experience. Am JMens Health
2014. pii: 1557988314539000 [Epub ahead of print].
34. Hoibraaten E, Qvigstad E, Anderson TO, Mowinckel MC,
Sandset PM. The effects of hormone replacement therapy
(HRT) on hemostatic variables in women with previous venous
thromboembolism—results from a randomized, double-blind,
clinical trial. Thromb Haemost 2001;85:775–881.
35. Canonico M, Plu-Bureau G, Lowe GD, Scarabin PY. Hormone
replacement therapy and risk of venous thromboembolism in
postmenopausal women: systematic review and meta-analysis.
36. Sinnesael M, Boonen S, Claessens F, Gielen E,
Vanderschueren D. Testosterone and the male skeleton: a dual
mode of action. J Osteoporos 2011;2011:240328.
37. Eisenberger A, Westhoff C. Hormone replacement therapy and
venous thromboembolism. J Steroid Biochem Mol Biol 2014;
38. FDA. Testosterone products: FDA/CDER statement-risk of
venous blood clots. FDA Medwatch, 2014.
39. Health Canada. Summary safety review. Testosterone replace-
ment products and cardiovascular risk. 2014.
40. Kobberoe Sogaard K, Schmidt M, Pedersen L, Horvath-Puho E,
Toft Sorensen H. 30-year mortality after venous thromboembo-
lism: a population-based cohort study. Circulation 2014;130:
41. Kim ED, Crosnoe L, Bar-Chama N, Khera M, Lipshultz LI. The
treatment of hypogonadism in men of reproductive age. Fertil
42. Cosmi B, Legnani C, Bernardi F, Coccheri S, Palareti G. Value of
family history in identifying women at risk of venous thromboem-
bolism during oral contraception: observational study. BMJ 2001;
43. Schambeck CM, Schwender S, Haubitz I, Geisen UE,
Grossmann RE, Keller F. Selective screening for the Factor V Lei-
den mutation: is it advisable prior to the prescription of oral
contraceptives? Thromb Haemost 1997;78:1480–3.
44. Aznar J, Vaya A, Estelles A, et al. Risk of venous thrombosis in
carriers of the prothrombin G20210Avariant and factor V Leiden
and their interaction with oral contraceptives. Haematologica
45. den HeijerM, Koster T, BlomHJ, et al. Hyperhomocysteinemia as
a risk factor for deep-vein thrombosis. N Engl J Med 1996;334:
46. Brandt JT, Triplett DA, Alving B, Scharrer I. Criteria for the diag-
nosis of lupus anticoagulants: an update. On behalf of the Sub-
committee on Lupus Anticoagulant/Antiphospholipid Antibody
of the Scientific and Standardisation Committee of the ISTH.
Thromb Haemost 1995;74:1185–90.
47. Bick RL. Antiphospholipid thrombosis syndromes. Clin Appl
Thromb Hemost 2001;7:241–58.
48. Greco TP, Amos MD, Conti-Kelly AM, Naranjo JD, Ijdo JW.
Testing for the antiphospholipid syndrome: importance of IgA
anti-beta 2-glycoprotein I. Lupus 2000;9:33–41.
49. GmbH RD. Testosterone II, estradiol II: electrochemilumines-
cence immunoassay, Reference #s 05200067 160, 03000079
50. Wang J, Sim AS, Wang XL, et al. Relations between markers of
renal function, coronary risk factors and theoccurrenceand severity
of coronary artery disease. Atherosclerosis 2008;197:853–9.
51. Al-Mondhiry H, Manni A, Owen J, Gordon R. Hemostatic effects
of hormonal stimulation in patients with metastatic prostate can-
cer. Am J Hematol 1988;28:141–5.
52. Kamphuisen PW, Eikenboom JC, Bertina RM. Elevated factor
VIII levels and the risk of thrombosis. Arterioscler Thromb
Vasc Biol 2001;21:731–8.
53. den Heijer M. Hyperhomocysteinaemia as a risk factor for venous
thrombosis: an update of the current evidence. Clin Chem Lab
54. den Heijer M, Willems HP, Blom HJ, et al. Homocysteine
lowering by B vitamins and the secondary prevention of
deep vein thrombosis and pulmonary embolism: a random-
ized, placebo-controlled, double-blind trial. Blood 2007;109:
55. Cattaneo M. Hyperhomocysteinemia and venous thromboembo-
lism. Semin Thromb Hemost 2006;32:716–23.
56. Glueck CJ, Hutchins RK, Jurantee J, Khan Z, Wang P. Thrombo-
philia and retinal vascular occlusion. Clin Ophthalmol 2012;6:
57. Gadelha T, Roldan V, Lecumberri R, et al. Clinical characteristics
of patients with factor V Leiden or prothrombin G20210A and a
first episode of venous thromboembolism. Findings from the
RIETE Registry. Thromb Res 2010;126:283–6.
Translational Research548 Freedman et al May 2015
58. Ivanov P, Komsa-Penkova R, Kovacheva K, et al. Impact of
thrombophilic genetic factors on pulmonary embolism: early
onset and recurrent incidences. Lung 2008;186:27–36.
59. Caprini JA, Goldshteyn S, Glase CJ, Hathaway K. Thrombophilia
testing in patients with venous thrombosis. Eur J Vasc Endovasc
60. Ben Salah R, Frikha F, Kaddour N, et al. Risk factor for deep
venous thrombosis in internal medicine: a retrospective study of
318 cases. Ann Cardiol Angeiol (Paris) 2014;63:11–6.
61. Turan O, Undar B, Gunay T, Akkoclu A. Investigation of inherited
thrombophilias in patients with pulmonary embolism. Blood Coa-
gul Fibrinolysis 2013;24:140–9.
62. Douketis JD, Julian JA, Crowther MA, et al. The effect of
prothrombotic blood abnormalities on risk of deep vein throm-
bosis in users of hormone replacement therapy: a prospective
case-control study. Clin Appl Thromb Hemost 2011;17:
63. Cherrier MM, Matsumoto AM, Amory JK, et al. The role of
aromatization in testosterone supplementation: effects on cogni-
tion in older men. Neurology 2005;64:290–6.
64. WuO, Robertson L, Twaddle S, et al. Screening for thrombophilia
in high-risk situations: a meta-analysis and cost-effectiveness
analysis. Br J Haematol 2005;131:80–90.
65. WuO, Robertson L, Twaddle S, et al. Screening for thrombophilia
in high-risk situations: systematic review and cost-effectiveness
analysis. The Thrombosis: Risk and Economic Assessment of
Thrombophilia Screening (TREATS) study. Health Technol
66. Blanco-Molina A, Monreal M. Venous thromboembolism in
women taking hormonal contraceptives. Expert Rev Cardiovasc
67. Beral V, Banks E, Reeves G. Evidence from randomised trials on
the long-term effects of hormone replacement therapy. Lancet