Venographic comparison of subcutaneous low-molecular weight heparin with oral anticoagulant therapy in the long-term treatment of deep venous thrombosis

Venog aphic compa ison o subcu aneous low–molecula weigh hepa in wi h o al an icoagulan he apy in he long- e m ea men o deep enous h ombosis Jose A. Gonzalez-Faja do, MD, Emilio A eba, MD, Ja ie Cas odeza, MD, Jose L. Pe ez, MD, Leopold Fe nandez, MD, Ignacio Agundez, MD, An onio M. Ma eo, MD, San iago Ca e a, MD, Vicen e Gu ié ez, MD, and Ca los Vaque o, MD, Valladolid, Spain Pu pose: The p ima y objec i e o his s udy was o e alua e wi h enog aphy he a e o h ombus eg ession a e a ixed dose o low–molecula weigh hepa in (LMWH) pe day o 3 mon hs compa ed wi h o al an icoagulan he apy o deep enous h ombo- sis (DVT). Seconda y endpoin s we e he compa isons o he e icacy and sa e y o bo h ea men s. Me hods: This s udy was designed as an open andomized clinical s udy in a uni e si y hos- pi al se ing. O he 165 pa ien s inally en olled in he s udy, 85 we e assigned LMWH he apy and 80 we e assigned o al an icoagulan he apy. In he g oup andomized o o al an icoagulan he apy, he pa ien s i s unde wen ea men in he hospi al wi h s an- da d un ac iona ed hepa in and hen couma in o 3 mon hs. Doses we e adjus ed wi h labo a o y moni o ing o main ain he in e na ional no malized a io be ween 2.0 and 3.0. Pa ien s in he LMWH g oup we e adminis e ed subcu aneous injec ions o ixed doses o 40 mg enoxapa in (4000 an i-Xa uni s) e e y 12 hou s o 7 days, and a e dis- cha ge om he hospi al, hey we e adminis e ed 40 mg enoxapa in once daily a ixed doses o 3 mon hs wi hou a labo a o y con ol assay. A quan i a i e enog aphic sco e (Ma de sco e) was used o assess he ex en o he enous h ombosis, wi h 0 poin s indi- ca ing no DVT and 40 poin s indica ing o al occlusion o all deep eins. The a e o h ombus educ ion was de ined as he di e ence in quan i a i e enog aphic sco es a e e mina ion o LMWH o couma in he apy as compa ed wi h he sco es ob ained on he ini ial enog aphic esul s. The e icacy was de ined as he abili y o p e en symp oma ic ex ension o ecu ence o enous h omboembolism (documen ed wi h enog ams o se ial lung scans). The sa e y was de ined as he occu ence o hemo hages. Resul s: A e 3 mon hs o ea men , he mean Ma de sco e was signi ican ly dec eased in bo h g oups in compa ison wi h he baseline sco e, al hough he e ec o he apy was signi ican ly be e a e LMWH he apy (49.4% educ ion) han a e couma in he a- py (24.5% educ ion; P< .001). LMWH he apy and male gende we e independen ly associa ed wi h an enhanced esolu ion o he h ombus. A lowe equency o symp o- ma ic ecu en enous h omboembolism was also shown in pa ien s who unde wen ea men wi h LMWH he apy (9.5%) han wi h o al an icoagulan he apy (23.7%; P < .05), al hough his di e ence was en i ely a esul o ecu ence o DVT. Bleeding complica ions we e signi ican ly ewe in he LMWH g oup han in he couma in g oup (1.1% s 10%; P< .05). This di e ence was caused by mino hemo hages. Couma in he apy and cance we e independen ly associa ed wi h an enhanced isk o complica- ions. Subcu aneous hepa in he apy was well ole a ed by all pa ien s. Conclusion: The pa ien s who we e alloca ed o unde go enoxapa in he apy had a sig- ni ican ly g ea e imp o emen in hei quan i a i e enog aphic sco e, a signi ican ly 283 F om he Di ision o Vascula Su ge y, Angio adiology (D A eba), and Epidemiology Depa men (D Cas odeza), Hospi al Clinico Uni e si a io. Rep in eques s: Jose A. Gonzalez-Faja do, MD, Di ision o Vascula Su ge y, Hospi al Clinico Uni e si a io, E-47011, Valladolid, Spain. Copy igh © 1999 by he Socie y o Vascula Su ge y and In e na ional Socie y o Ca dio ascula Su ge y, No h Ame ican Chap e . 0741-5214/99/$8.00 + 0 24/1/97709 Low–molecula weigh hepa in (LMWH) has been shown o be a leas as e ec i e and sa e as in a- enous adjus ed-dose hepa in in he ini ial ea men o p oximal deep ein h ombosis (DVT).1-10 Howe e , common p ac ice is o ca y ou ea men wi h o al an icoagulan s o a leas 3 mon hs as a o m o seconda y p e en ion.11 O al an icoagulan he apy is associa ed wi h a signi ican isk o bleeding compli- ca ions, and pa ien s who unde go such he apy equi e equen labo a o y moni o ing. Some o hese pa ien s ha e ela i e o absolu e con aindica ions o he use o o al an icoagulan s.12 Unde such ci cum- s ances, an al e na i e o p e en delayed h omboem- bolic ecu ence is adjus ed dose o subcu aneous hepa in wice daily.11-13 Compa ed wi h un ac iona - ed hepa in, LMWH has a longe hal li e, signi ican ly ewe hemo hagic complica ions, and a g ea e bioa ailabili y a low doses, which allows once-a-day schedule and he e o e may be p e e able o s anda d hepa in especially o long- e m adminis a ion.14 Some s udies ha e ecen ly e alua ed he ole o LMWH as an al e na i e o o al an icoagula ion in he p e en ion o ecu en h omboembolism.12,15,16 They sugges ha ixed doses o LMWH appea o be qui e e ec i e and sa e, bu limi ed da a on he basis o enog aphic obse a ions a e a ailable. P e ious in es- iga ions ha compa ed LMWH wi h s anda d un ac- ioned hepa in in he ini ial ea men o DVT e ealed a phlebog aphic imp o emen in he pa ien s who unde wen ea men wi h LMWH when he enog ams we e pe o med app oxima ely 1 o 2 weeks la e ,14 bu no signi ican di e ences we e seen a he 6 mon h ollow-up examina ion.17 F om he exis ing e idence, i is likely ha a ma ked o o al educ ion o h ombi will educe he incidence o pos - h ombo ic synd ome.18 Because a longe du a ion o he ea men wi h LMWH may inc ease he success a e o ecanaliza ion o he occluded eins, i seems legi ima e o use phlebog aphic endpoin s.19 Fo his eason, he p ima y objec i e o he p esen s udy was o e alua e wi h enog aphy he a e o h ombus eg ession a e a single subcu aneous injec ion o LMWH (enoxapa in, 40 mg) pe day o 3 mon hs as compa ed wi h o al an icoagulan he apy o he ea men o pa ien s wi h DVT. Seconda y endpoin s we e o compa e he e icacy (p e en ion o symp o- ma ic enous h omboembolism) and sa e y (occu - ence o hemo hage) o bo h ea men s. MATERIAL AND METHODS S udy design. This was an open, andomized s udy compa ing con en ional couma in he apy wi h a 3-mon h cou se o enoxapa in (4000 an i-Xa uni s; 40 mg) subcu aneously once daily in a ixed dose. The p ima y endpoin was he abili y o eopen h omboses eins, de ined as he di e ence in quan i a i e enog aphic sco es a e e mina ion o LMWH o couma in he apy compa ed wi h he sco es ob ained on he ini ial enog aphy. The sec- onda y endpoin s we e he de elopmen o symp o- ma ic ecu en pulmona y embolism o symp o- ma ic ecu en enous h ombosis (documen ed wi h se ial pe usion lung scans o enog ams) and bleeding episodes. Randomiza ion was achie ed by means o a p esc ibed schedule. In o med w i en o e bal consen was ob ained om all pa ien s in he s udy. The s udy p o ocol was accep ed by he Ins i u ional Re iew Boa d. Pa ien s. Consecu i e eligible pa ien s wi h clinically suspec ed DVT con i med wi h con as enog aphy we e en olled in he s udy acco ding o a compu e schedule. The easons o exclusion we e: clinically suspec ed pulmona y embolism, cu - en ly ac i e bleeding o coagula ion abno mali y o diso de s con aindica ing an icoagulan he apy, p egnancy, wo o mo e p e iously documen ed episodes o DVT o pulmona y embolism, ongoing an icoagulan ea men a he ime o e e al, his- o y o hepa in-induced h ombocy openia, ca al il- e inse ed, alle gic eac ion o con as ma e ial, a coagula ion-inhibi o de iciency, a lupus an icoagu- lan , o an iphospholipid an ibodies. In each pa ien , an icoagulan he apy was s a ed as soon as possible a e DVT had been documen ed objec i ely wi h ascending con as enog aphy. Whene e possible, he pa ien s we e allowed o walk on he hi d day o ea men , wea ing elas ic comp essi e s ockings. Regimens. In he pa ien s andomized o o al an icoagulan he apy, an in a enous bolus o 100 U/kg o un ac ioned hepa in was adminis e ed, ol- lowed by a con inuous in a enous in usion o hepa in. The ac i a ed pa ial h omboplas in ime JOURNAL OF VASCULAR SURGERY 284 Gonzalez-Faja do e al Augus 1999 lowe ecu ence a e o symp oma ic enous h omboembolism, and a signi ican ly lowe incidence o bleeding han pa ien s who unde wen ea men wi h couma in. LMWH can be used on an ou pa ien basis as a sa e and mo e e ec i e al e na i e o classical o al an icoagulan he apy o he seconda y p ophylaxis o selec ed pa ien s wi h DVT. (J Vasc Su g 1999;30:283-92.) was measu ed 4 hou s a e he beginning o in a- enous hepa in ea men , and he es was epea ed a in e als o 4 o 6 hou s un il he esul was wi h- in he p esc ibed he apeu ic ange ( a io, 1.5 o 2.0) du ing he ini ial 24 hou s o he apy. O al ea men wi h couma in (ini ial dose o 5 mg) was s a ed in pa ien s on day 5 o hepa in ea men . The couma in dose was adjus ed daily o main ain he in e na ional no malized a io (INR) be ween 2.0 and 3.0. Subsequen ly, hepa in ea men was discon inued and he couma in dose was adjus ed wi h labo a o y moni o ing o 3 mon hs. Al hough du ing his s udy he classical app oach o he ea - men o DVT was used, he esul s o wo andom- ized clinical ials20,21 ha e shown ha o al an ico- agulan he apy s a ing wi h hepa in a he ime o diagnosis is as e ec i e and sa e as he con en ional egimen. The in ensi y o an icoagula ion he apy in he i s 3 mon hs was exp essed as he pe cen age o ime du ing which a pa ien had a speci ic INR (<2.0, 2.0 o 3.0, o >3.0), wi h his pe iod calcu- la ed wi h linea in e pola ion. No pa ien s unde - wen o al an icoagulan he apy a e 3 mon hs o ea men , unless hey had symp oma ic ecu en h omboembolism. Pa ien s in he LMWH g oup we e adminis e ed subcu aneous injec ions e e y 12 hou s o ixed doses o 40 mg enoxapa in (co e- sponding o 4000 in e na ional ac o Xa inhibi o y uni s) o 7 days. A e discha ge om he hospi al, he pa ien s unde wen ea men wi h 40 mg enoxapa in subcu aneously once daily, which was usually adminis e ed by he pa ien s hemsel es o by ela i es and only occasionally by nu ses. LMWH was adminis e ed a ixed doses, wi hou a labo a o- y con ol assay o 3 mon hs. Main ou come measu es. Venog aphy was pe - o med wi h long-leg ilms and no ionic con as ma e ial acco ding o he me hod o Rabino and Paulin.22 The c i e ia o DVT we e an in aluminal illing de ec con i med in a leas wo di e en p o- jec ions and no illing o a enous de ec despi e epea ed injec ions wi h con as ma e ial. A quan i- a i e enog aphic sco e (Ma de sco e23) was used o assess he ex en o he enous h ombosis, wi h 0 poin s indica ing no DVT and 40 poin s indica ing o al occlusion o all deep eins (Table I). Con ol phlebog aphies we e pe o med ou inely in all pa ien s a e 3 mon hs o ea men . Each pa ien was assigned o one o he ollowing g oups on he basis o di e ences on he Ma de sco e be ween ini- ial and pos - ea men phlebog aphy: inc eased h ombosis (an inc ease in sco e poin s), unchanged (unal e ed o al sco e o ≤10% dec ease in sco e poin s), pa ly clea ed (>10% o ≤50% dec ease in sco e poin s), subs an ially clea ed (>50% o 89% dec ease in sco e poin s), and comple ely lysed (≥90% dec ease in sco e poin s). Pe usion lung scanning and ches adiog aphy we e pe o med o all pa ien s a baseline. To exclude bias, he assess- men o enog ams and lung scans we e sco ed by wo obse e s who we e blind o ea men alloca- ion and o he sequence in which he es s we e done (be o e o a e ea men ). The e icacy o ea men was de ined as he abili y o p e en symp- oma ic ex ension o ecu ence o enous h om- boembolism. The sa e y was de ined as he occu - ence o hemo hages. Bleeding was de ined as majo bleeding i i was in ac aneal o e ope i- oneal o i i p oduced a dec ease in he hemoglo- bin le el o a leas 2.0 g/dL, su icien o necessi- a e discon inua ion o ea men o he ans usion o 2 o mo e uni s o blood. Bleeding was de ined as mino bleeding i i did no mee he c i e ia o majo bleeding. Su eillance and ollow-up. A e discha ge, all he pa ien s we e seen in ou ascula clinic e e y mon h du ing he i s imes e and hen e e y 3 mon hs du ing a o al o 1 yea o ollow-up. They we e ins uc ed o come o he hospi al immedia e- ly i symp oms o signs o ecu en DVT, pul- mona y embolism, o bleeding de eloped. Those pa ien s wi h suspec ed ecu en DVT unde wen con as enog aphy. Recu en DVT was de ined as a cons an in aluminal illing de ec no p esen he i s day. Pa ien s wi h clinically suspec ed pul- mona y embolism unde wen ano he pe usion lung scan and ches adiog aphy. The diagnosis was made on he basis o he p esence o a leas one seg- men al de ec no seen on he p eceding scan and no abno mali y on he ches adiog aph a ea. I he esul s we e inconclusi e, pulmona y angiog aphy JOURNAL OF VASCULAR SURGERY Volume 30, Numbe 2 Gonzalez-Faja do e al 285 Table I. Venog aphic quan i a ion o h ombosis (Ma de sco e) Deep eins Sco e* Iliac 6 Common emo al 4 Supe icial emo al 10 Popli eal 4 An e io ibial 4 (2 each) Pos e io ibial 6 (3 each) Pe oneal 6 (3 each) *To al occlusion o non illing o a gi en ein was assigned he maximum sco e; segmen al occlusion o illings de ec s we e gi en lesse sco es in p opo ion o he deg ee o in ol emen . was pe o med. Du ing he 3 mon hs o ea men , pla ele coun s we e ob ained each mon h in he pa ien s unde going LMWH he apy o ule ou he possibili y o hepa in-induced h ombocy openia. Adhe ence o he s udy egimens was moni o ed by e iews o he pa ien s’ cha s. A each isi , pa ien s we e ques ioned abou he appea ance o new symp- oms, bleeding, swelling, hea iness o leg, leg i ed- ness, o pain. The clinical s a us a e 3 mon hs o ea men was g aded subjec i ely by each pa ien as imp o ed, unchanged, o wo se in ela ion o he p e ea men s a us. S a is ical analysis. On he basis o esul s om p e ious ials,24 he p opo ion o pa ien s wi h an unchanged o imp o ed Ma de sco e a e LMWH ea men was assumed o be 90%. By sample size, we de e mined a 15% absolu e di e ence in he p opo ions o pa ien s whose condi ions we e unchanged and imp o ed ha we e needed o show a s a is ically signi ican di e ence be ween he wo ea men s. A an 80% powe o showing his di - e ence a he 5% signi icance le el, wi h a one-sided es , a o al o 78 pa ien s would be necessa y a leas in each ea men g oup. The numbe o pa ien s planned o inclusion in o each g oup was he e o e 92, o accoun o an an icipa ed d op-ou a e o 15%. Quan i a i e da a we e exp essed as mean ± s anda d e o . Con idence in e als (CI) o 95% o he di e ence be ween he wo ea men g oups we e calcula ed wi h he no mal app oxima ion o he binomial dis ibu ion. The a es o asymp oma ic pulmona y embolism, ecu en h omboembolism, bleeding, dea h, deg ee o h ombus eg ession, and clinical s a us in he wo g oups we e compa ed by means o Fishe exac es o χ2 es as app op ia e. The changes in enog aphic sco e we e analyzed wi hin and be ween g oups wi h wo-sample and ma ched-pai and unpai ed es s. Uni a ia e analy- sis was pe o med o iden i y ac o s ha a ec ed he di e ences on he Ma de sco e be ween ini ial and pos - ea men phlebog aphy. A mul i a ia e s ep- wise eg ession model was used o iden i y indepen- den a iables ha could in luence he pe cen age o change in Ma de sco e and complica ions o bo h g oups. Two-sided P alues o less han .05 we e conside ed signi ican . So wa e JMP 3.1 om he S.A.S. Ins i u e Inc (Ca y, NC) was used. RESULTS Pa ien s. F om June 1994 o June 1997, 257 consecu i e pa ien s wi h clinically suspec ed DVT unde wen ea men in ou hospi al. Eligible JOURNAL OF VASCULAR SURGERY 286 Gonzalez-Faja do e al Augus 1999 Fig 1. Venog aphic changes wi h 40 mg enoxapa in (4000 an i-Xa uni s). A, To al occlusion o popli eal and emo al eins in pa ien wi h deep enous h ombosis. B, Comple e ecana- liza ion a e ixed dose pe day o 3 mon hs o ea men . AB pa ien s (n = 185) had DVT con i med wi h con as enog aphy, pe usion lung scanning wi hin 48 hou s o s udy en y, and ches adiog aphy. O hese, 93 we e assigned o unde go LMWH he apy and 92 o unde go ea men wi h couma in. Twen y pa ien s we e excluded om he analysis (eigh in he LMWH g oup, and 12 in he couma in g oup) because he second enog am was no ob ained (n = 12), he egimen o ea men was no pe o med co ec ly by he pa ien (n = 5), o he pa ien s we e los du ing he ollow-up pe iod (n = 3). Fo he 165 pa ien s inally en olled in o he s udy, he baseline clinical cha ac e is ics a e shown in he Table II. The ea men g oups we e compa- able a en y excep o age (younge in he couma in g oup) and incidence o silen pulmona y embolism (highe in he couma in g oup). To assess he possible e ec o his po en ial age and asymp o- ma ic pulmona y embolism imbalance, mul iple logis ic eg ession was used. No signi ican e ec was ound. Deg ee o h ombus eg ession. The in ensi y o o al an icoagulan he apy in he couma in g oup was 15% wi h INR less han 2.0, 64% wi h INR om 2.0 o 3.0, and 21% wi h INR mo e han 3.0. A sec- ond enog am was pe o med in all pa ien s. The changes in he ex en o enous h ombosis on enog- aphy a e summa ized in Table III. No signi ican di - e ence in Ma de sco e was obse ed be ween he wo ea men g oups a inclusion. A e 3 mon hs o ea men , he mean Ma de sco e was signi ican ly dec eased in bo h g oups in compa ison wi h he baseline sco e, al hough he e ec o he apy was sig- ni ican ly be e a e LMWH he apy (49.4% educ- ion o he Ma de sco e) han a e couma in he apy (24.5% educ ion o he Ma de sco e; P< .001; Fig 1). The esul s o he s a i ied analysis o he e olu- ion o he h ombus (Table III), on he basis o he pe cen age o he di e ences o sco e be ween he ini- ial and pos - ea men phlebog aphy, also showed a s a is ically signi ican supe io i y o LMWH he apy o e couma in he apy (P< .001). JOURNAL OF VASCULAR SURGERY Volume 30, Numbe 2 Gonzalez-Faja do e al 287 Table II. Clinical cha ac e is ics o pa ien s wi h deep enous h ombosis ea ed wi h enoxapa in o couma in Enoxapa in (n = 85) Couma in (n = 80) P alue Mean age (yea s) 62.7 ( , 19 o 83) 58.3 ( , 20 o 82) <.05 Sex (male: emale) 41:44 46:34 Days since onse o symp oms 4.9 ( , 1 o 30) 4.1 ( , 1 o 20) Asymp oma ic pulmona y embolism 21 (24.7%) 45 (56.2%) <.001 Risk ac o s o DVT Recen auma, su ge y, o immobiliza ion 34 (40%) 23 (28.77%) Cance 10 (11.7%) 8 (10%) Unknown 41 (48.2%) 49 (61.2%) Loca ion o h ombus Cal wi h popli eal ein 6 (7.05%) 7 (8.7%) Popli eal and emo al ein 43 (50.5%) 35 (43.7%) Popli eal, emo al, and iliac ein 26 (30.5%) 30 (37.5%) Iliac ein 10 (11.7%) 8 (10%) DVT, Deep enous h ombosis. Table III. Venog aphic e alua ion Enoxapa in (n = 84)* Couma in (n = 80) P alue Ma de sco e Be o e ea men 24.7 ± 1.1 26.06 ± 1.1 A e ea men 12.5 ± 1.05 19.7 ± 1.1 <.001 Di e ence (∆) –12.2 ± 0.9 –6.4 ± 0.8 <.001 Fa e o he h ombus <.001 Inc ease in size 2 (2.3%) 8 (10%) No change 3 (3.5%) 13 (16.2%) Pa ial clea ance 39 (46.4%) 45 (56.2%) Subs an ial clea ance 24 (28.5%) 13 (16.2%) Comple e lysis 16 (19.04%) 1 (1.2%) *One pa ien died o massi e pulmona y embolism. Uni a ia e analysis shows ha no signi ican di - e ences we e obse ed in he dis ibu ion o changes in enog aphic sco e be ween bo h g oups when he cause o DVT was cance o he h ombus was placed only a genicula e o iliac le el (Table IV). In con as , a highe eopening a e o h om- bosed eins was shown in he pa ien s assigned o unde go LMWH he apy as compa ed wi h he pa ien s assigned o unde go ea men wi h couma in when he cause o DVT was ecen au- ma, su ge y, o immobiliza ion (P= .0037), when he DVT was idiopa hic (P= .0009), o when he h ombus was placed in popli eal and emo al eins (P= .0011) o a ec ed comple ely all o he limb (P = .01005). In a s ep-wise linea eg ession model, LMWH he apy (P< .0001) and male gende (P= .0199) we e independen ly associa ed wi h an enhanced esolu ion o he h ombus (Table V). The 2 alue was 0.11435. Recu en h omboembolism. Symp oma ic ex ension o ecu en enous h omboembolism con i med wi h objec i e es s de eloped in eigh pa ien s (9.5%) o he LMWH g oup and in 19 pa ien s (23.7%) assigned o unde go couma in he apy (P= .0196; Table V). O he 19 e en s in he couma in g oup, 15 in ol ed ecu en h om- bosis (13 ecu ences we e in he same limb, wo we e con ala e al) and ou in ol ed pulmona y embolism ( om which all pa ien s equi ed ca a il- e placemen ). O he eigh e en s in he LMWH g oup, i e in ol ed ecu en h ombosis (all in he same limb) and h ee in ol ed pulmona y embolism ( om which one pa ien died on he i h day a e andomiza ion). The absolu e di e ence in he e- quency o pulmona y embolism did no each s a is- ical signi icance (P= .7149). Howe e , signi ican di e ences we e obse ed in he equency o ecu - en h ombosis (P= .0161). Du ing he 12-mon h su eillance pe iod, wo pa ien s in he couma in g oup had a ecu ence o symp oma ic DVT docu- men ed wi h enog aphy a e discon inua ion o he an icoagulan ea men (one in he 20 h week, and he o he in he 24 h week). In he LMWH g oup, h ee pa ien s had symp oma ic ecu en DVT documen ed wi h enog aphy (one in he 20 h week, ano he in he 24 h week, and ano he in he 32nd week). Only one pa ien o he couma in g oup was eadmi ed o he hospi al in he 21s week because o symp oma ic ecu en pulmona y embolism con i med wi h pe usion lung scanning and ches adiog aphy. No symp oma ic pulmona y embolisms we e obse ed in he LMWH g oup. Recu en h omboembolism was associa ed wi h ecen su ge y o auma in h ee pa ien s ( h ee JOURNAL OF VASCULAR SURGERY 288 Gonzalez-Faja do e al Augus 1999 Table IV. Uni a ia e analysis o isk ac o s and loca ion o h ombus associa ed wi h di e ences on he Ma de sco e be ween ini ial and pos - ea men phlebog aphy Enoxapa in ∆Ma de sco e Couma in ∆Ma de sco e P alue Risk ac o s o DVT Recen auma, su ge y, o immobiliza ion 11.4 ± 1.3 6.2 ± 1.04 <.01 Cance 8.1 ± 2.1 3.5 ± 1.5 Unknown 13.7 ± 1.5 7 ± 1.2 <.001 Loca ion o h ombus Cal wi h popli eal ein 7 ± 2.3 5.1 ± 2.2 Popli eal and emo al ein 11.8 ± 1.3 6.1 ± 1.06 <.01 Popli eal, emo al, and iliac ein 15.2 ± 2.07 8.7 ± 1.2 <.05 Iliac ein 8.3 ± 1.8 0.25 ± 4.1 ∆ Ma de sco e, Ma de sco e be o e ea men – Ma de sco e a e ea men ; DVT, deep enous h ombosis. Table V. Independen de e minan s o he pe cen age o change in Ma de sco e as selec ed wi h mul i- a ia e s ep-wise linea eg ession model Signi ican independen 95% CI a iables Lowe Uppe F alue P alue T ea men (enoxapa in) –28.329 –9.889 16.5 .0001 Sex (male) –20.348 –1.851 5.53 .0199 F, Fishe -Snedeco . R2= 0.11435. cases in he LMWH g oup, and none in he couma in g oup), cance in 10 pa ien s ( i e cases in he LMWH g oup, and i e in he couma in g oup), and idiopa hic disease in 20 pa ien s ( h ee cases in he LMWH g oup, and 17 in he couma in g oup). Bleeding complica ions. Bleeding complica- ions we e signi ican ly ewe in he LMWH g oup (1.1% s 10%; P= .0160; Table VI). Majo bleeding occu ed du ing o immedia ely a e he ini ial he - apy in one pa ien unde going LMWH he apy (1.1%) and in 2 pa ien s unde going ea men wi h couma in (2.5%; P= .6136). The hemo hage in he LMWH g oup consis ed o uppe gas oin es inal bleeding, and in he couma in g oup, he e we e one uppe gas oin es inal bleeding and one e ope i oneal bleeding wi h hema u ia. Mino hemo hagic complica ions occu ed in none o he pa ien s who unde wen LMWH he apy and in six pa ien s who unde wen ea men wi h couma in (7.5%; P= .0121). The mino bleedings in he couma in g oup consis ed o wo hema u iax, h ee epis axis, and one hemop ysis. Small hema omas in he abdominal wall we e seen in pa ien s assigned o unde go LMWH he apy. In a s ep-wise logis ic eg ession model (Table VII), couma in he apy (odds a io, 2.82; CI, 1.719 o 4.623; P< .0001) and cance (odds a io, 2.15; CI, 1.101 o 4.233; P = .025) we e independen ly associa ed wi h an enhanced isk o complica ions (symp oma ic ecu - en h omboembolism and bleeding). These con- clusions emained unchanged when he ew pa ien s wi h documen ed DVT who we e wi hd awn a e andomiza ion because o p o ocol iola ion we e included in an in en ion- o- ea analysis. Dea hs. One pa ien om he LMWH g oup died o massi e pulmona y embolism, con i med wi h angiog aphy, on he i h day a e he ini ial ea men . Du ing he 12-mon h s udy pe iod, ou pa ien s who we e assigned o he LMWH g oup died, as compa ed wi h h ee pa ien s assigned o he couma in g oup. The causes o dea h included can- ce ( i e pa ien s) and ca dio ascula disease ( wo pa ien s). Compliance. Only one case o h ombocy ope- nia (<100.000/mm3) was obse ed immedia ely a e he ini ial he apy wi h un ac ioned hepa in. This diso de dissapea ed wi h o al an icoagulan he apy. Subcu aneous hepa in he apy was well ol- e a ed by all pa ien s. No pa ien s showed hepa in- induced h ombocy openia in he 3-mon h g oup wi h LMWH. The clinical s a us g aded subjec i ely by each pa ien was signi ican ly be e a e LMWH JOURNAL OF VASCULAR SURGERY Volume 30, Numbe 2 Gonzalez-Faja do e al 289 Table VI. Th omboembolic and bleeding complica ions du ing ini ial ea men and ollow-up examina ion Enoxapa in (n = 85) Couma in (n = 80) P alue T ea men pe iod (3 mon hs) Majo bleeding 1 2 Mino bleeding 0 6 .0121 To al e en s 1 8 .0160 Recu ence o DVT 5 15 .0161 Pulmona y embolism 3* 4 To al e en s 8 19 .0196 Su eillance pe iod (9 mon hs) Recu ence o DVT 3 2 Pulmona y embolism 0 1 To al e en s 3 3 DVT, Deep enous h ombosis. *One a al pulmona y embolism. Table VII. Independen de e minan s o complica ions ( enous h omboembolism and bleeding) as selec ed wi h mul i a ia e s ep-wise logis ic eg ession model Signi ican independen 95% CI a iables Lowe Uppe Odds a io P alue T ea men (couma in) 1.719 4.623 2.82 .0001 Risk ac o (cance ) 1.101 4.233 2.15 .0250 CI, Con idence in e al. he apy han a e couma in he apy (P< .001). The e was an imp o emen a e 3 mon hs o ea - men in ela ion o he p e ea men s a us in 68 o 84 pa ien s (80.9%) in he LMWH g oup and in 40 o 80 (50%) in he couma in g oup. The e we e no changes in 15 o 84 pa ien s (17.8%) in he LMWH g oup and in 35 o 80 (43.7%) in he couma in g oup. Finally, he e was wo sening in one o 84 pa ien s (1.1%) and i e o 80 (6.2%), espec i ely. DISCUSSION This s udy showed ha enoxapa in, a LMWH, no only can be used sa ely and e ec i ely o ea DVT a home o long e m bu also ha i s abili y o eopen h omboses eins is be e han couma in, an an i– i - amin K o al an icoagulan . Al hough he mean Ma de sco e was signi ican ly imp o ed in bo h g oups a he 3-mon h ollow-up examina ion, a highe deg ee o ecanaliza ion was shown wi h enog aphy in he pa ien s assigned o unde go LMWH he apy (49.4%) as compa ed wi h he pa ien s assigned o unde go ea men wi h couma in (24.4%). Mo eo e , in a mul i a ia e analysis, he male pa ien s we e indepen- den ly associa ed wi h an enhanced esolu ion o he h ombus. A h ombus educ ion o a leas 30% o mo e is a p esen used in se e al ials as a sign o clin- ical bene i o he indi idual pa ien ,19 wi h he expec ancy ha a majo ecanaliza ion may also esul in a educed incidence o pos - h ombo ic syn- d ome.18,19 Because o his, e idence on he basis o enog aphic indings sugges s ha a LMWH may educe he isk o he la e sequelae o DVT, al hough i will open o discussion o wha ex en he educ ion o h ombus size eally is accomplished by long- e m bene i s. The 12-mon h ollow-up pe iod used in his s udy is oo sho o e lec he ue de elopmen o clinical pos - h ombo ic synd ome. Howe e , LMWH was subjec i ely associa ed by each pa ien wi h be e clinical s a us han couma in he apy a e 3 mon hs o ea men . Long- e m enog aphic ollow-up da a in pa ien s wi h DVT a e limi ed.17 Ou s udy is he i s o analyse a longe du a ion o ea men wi h LMWH and he endency o h ombus eg ession in compa ison wi h pa ien s unde going o al an icoag- ulan he apy as a o m o seconda y p e en ion. The mechanism by which LMWH acili a es h ombus eg ession is no unde s ood. P obably he inhibi- ion o h ombin o ma ion, he lesse isk o pla ele agg ega ion, and local e ec s a he endo helial le el play a c i ical ole.25,26 Indeed, endo helial cells a e he p incipal physiologic sou ce o issue- ype plas- minogen ac i a o , and he subcu aneous adminis- a ion o LMWH causes an inc ease in plasma o is- sue- ype plasminogen ac i a o wi h peak a 3 hou s a e injec ion.27,28 The ecanaliza ion o h ombosed eins was qui e di e en in hese wo pa ien g oups acco ding o he cause o DVT and he localiza ion o he enous occlusion. Pa icula ly, he e was a signi ican end in a o o LMWH when he cause o DVT was ecen auma, su ge y, o immobiliza ion, when he DVT was idiopa hic, o when he h ombus was placed in popli eal and emo al eins o a ec ed com- ple ely all o he limb. By con as , i he cause o DVT was cance o he in aluminal illing de ec was exclusi ely placed a genicula e o iliac le el, no sig- ni ican di e ences we e obse ed be ween pa ien s assigned o unde go LMWH he apy and pa ien s assigned o unde go ea men wi h couma in. Ne e heless, because o he la ge s anda d e o and he ela i ely small numbe o pa ien s among hese subg oups, his s udy does no ha e su icien s a is- ical powe o de ec signi ican di e ences. Fu he s udies a e necessa y be o e de ini i e conclusions can be eached. LMWH ha e unde gone limi ed in es iga ions o he seconda y p e en ion o enous h omboem- bolism.9In hese s udies, pa ien s i s unde wen ea men in he hospi al wi h s anda d hepa in, and hen he apy wi h LMWH was compa ed wi h he a- py wi h wa a in15,16 o subcu aneous s anda d hepa in12 o 3 o 6 mon hs. Mon eal e al12 showed, in a consecu i e se ies o pa ien s wi h con aindica- ion o couma in he apy, ha pa ien s who unde - wen ea men wi h LMWH had a lowe equency o ei he ecu en pulmona y embolism o mino bleeding in compa ison wi h un ac ioned hepa in, al hough none o hese di e ences we e s a is ically signi ican pe haps because o he small numbe o pa ien s (be a e o ). Pini e al15 con i med hese indings in a su icien ly la ge andomized s udy o pa ien s wi h DVT, al hough he diagnosis o ecu - en DVT was o en made on he basis o nonin asi e es ing and diagnosis o pulmona y embolism could be ques ioned because baseline lung scans we e no pe o med. Despi e hese ac s, hey showed essen- ially no di e ence in he incidence o ecu en h ombosis be ween pa ien s who unde wen ea - men wi h LMWH (enoxapa in 40 mg) and pa ien s who unde wen ea men wi h wa a in, bu pa ien s alloca ed o unde go LMWH he apy had a signi i- can ly lowe incidence o bleeding. In he hi d ial, which has been epo ed in abs ac o m, Kakka 16 co obo a ed hese indings wi h dal epa in. In ou s udy, a es o symp oma ic ecu en JOURNAL OF VASCULAR SURGERY 290 Gonzalez-Faja do e al Augus 1999 h omboembolism and bleeding also ended o occu less in he pa ien s unde going enoxapa in he apy han in hose pa ien s unde going couma in he apy (odds a io, 2.82; CI, 1.719 o 4.623). As expec ed, a low bleeding a e was obse ed in he LMWH g oup. The egimen o 40 mg enoxapa in, a a ixed dose once daily subcu aneously, esul ed in 1.1% e - sus 10% o hemo hagic complica ions compa ed wi h he couma in ea men (P< .05). This di e ence was a esul o mino hemo hages. Concomi an ly, a lowe equency o symp oma ic ecu en enous h omboembolism was shown in pa ien s who unde - wen ea men wi h LMWH (9.5%) han wi h o al an icoagulan s (23.7%; P< .05), al hough his di e - ence was en i ely a esul o ecu ence o DVT. Thus, we ind a s a is ically signi ican supe io i y o LMWH o e couma in ega ding bo h e icay and sa e y. Al hough he doses ha we used in he p esen s udy we e ac ually p ophylac ic doses a he han ea men doses, especially o he i s 7 days o ea men , i is impo an o conside whe he ou esul s we e caused by a ue p ope y o enoxapa in because ade- qua e dosage inding s udies a e no a ailable.29,30 The ac ha a pa ien died in he LMWH g oup o a ecu en pulmona y embolism on he i h day o andomiza ion sugges s ha he dosage used may be insu icien and ha highe le els should be p e e able in hese pa ien s. Pa icula ly, we belie e ha an adjus ed dose pe kilog am o body weigh would be ecommendable o he managemen o DVT, al hough u he dose- ela ed s udies a e necessa y o de ine minimal and maximal doses. Because he wo egimens we e gi en by di e en ou e and because dose adjus men s we e necessa y in he couma in g oup, we could no use a double- blind design. To minimize bias in he assessmen o symp oma ic ecu en enous h omboembolism, all suspec ed ecu ences we e e alua ed wi h objec i e es s and he da a we e analyzed by wo blinded obse e s. Gi en ha we did ou inely sc een all pa ien s wi h enog aphy a 3 mon hs, silen enous ecu ences may no ha e gone unde ec ed, bu silen pulmona y embolism may ha e occu ed, because lung scans we e only aken o assis diagnosis o symp oma ic pulmona y embolism. Some conce n may a ise om he inding ha he e was a highe equency o silen pulmona y embolism in pa ien s alloca ed o he couma in g oup han in pa ien s allo- ca ed o he LMWH g oup, which could ha e po en- ially skewed he esul s. Howe e , a mul iple logis ic eg ession showed ha his a iable was no inde- penden ly associa ed wi h he ecu ence o enous h omboembolism and bleeding. In addi ion, bo h g oups we e compa able acco ding o loca ion o h ombus, p edisposing ac o s, and ex ension o he enous h ombosis. The highe equency o silen pulmona y embolism con i ms p e ious s udies ha show ha , by he ime he DVT has been diag- nosed, many pa ien s ha e al eady had a pulmona y embolism.31 The numbe o dea hs and, in pa icula , he p o- po ion o dea hs in pa ien s wi h cance ( h ee o i e in he LMWH g oup, and wo o h ee in he couma in g oup) we e compa able in bo h ea - men g oups. Ou da a he e o e do no subs an ia e he hypo hesis ha LMWH adminis a ion could exe a a o able e ec on cance p og ession.1,6 Howe e , he pa ien s wi h cance we e indepen- den ly associa ed wi h mo e complica ions (odds a io, 2.15; CI, 1.101 o 4.233). Rega ding he compliance o he pa ien s wi h LMWH he apy, his was well ole a ed and allowed he pa ien s o be ully ambulan . None o he pa ien s had nega i e eac ions o unde going ea - men a home wi h a subcu aneous injec ion daily o 3 mon hs, and no pa ien s showed hepa in- induced h ombocy openia in he 3-mon h g oup wi h LMWH he apy. Howe e , mos pa ien s had small hema omas in he injec ion si es. Al hough os eopo osis is a well-known side-e ec o hepa in he apy, pa icula ly in long- e m he apy, we ha e no pe o med any bone densi y measu emen s, x- ays, o o he s udies o look o his po en ial com- plica ion. Acco ding o he only s udy ha has ana- lyzed his ac , LMWH he apy was associa ed wi h a lesse isk o bone ac u e han un ac iona ed hepa in he apy.12 This means ha LMWH seems o be a good al e na i e o s anda d hepa in o pa ien s in whom o al an icoagulan he apy is con aindica - ed, especially elde ly pa ien s, in whom spine ac- u es a e qui e common. In conclusion, he pa ien s alloca ed o unde go ea men wi h enoxapa in had a signi ican ly g ea e imp o emen in hei clinical s a us and quan i a i e enog aphic sco e, a signi ican ly lowe ecu ence a e o symp oma ic enous h omboembolism, and a signi ican ly lowe incidence o bleeding han pa ien s who unde wen ea men wi h couma in. These p omising esul s ob ained wi h LMWH he apy adminis e ed by subcu aneous injec ion, wi hou lab- o a o y moni o ing, aise he possibilli y ha hese agen s can be used on an ou pa ien basis as a sa e and mo e e ec i e al e na i e o classic o al an icoag- ulan he apy o he seconda y p ophylaxis o selec - ed pa ien s wi h es ablished DVT. Pa icula ly good candida es would be p egnan women, pa ien s wi h a JOURNAL OF VASCULAR SURGERY Volume 30, Numbe 2 Gonzalez-Faja do e al 291