Thromboembolie veineuse Recherche de cancer André Roussin MD, FRCP Hôpital Notre-Dame Professeur adjoint Université de Montréal .org Physiopathologie Épidémiologie / littérature Synopsis Examens suggérés ACCP 2004 2004

Cancer et thrombose: physiopathologie Bick R L NEJM 2003349: 109-111

260 pts consécutifs⇒ 250 pts suivis x 2 ans “Deep-vein thrombosis and the incidence of subsequent symptomatic cancer” Prandoni P et al.NEJM 1992; 327: 1128-1133 260 pts consécutifs⇒ 250 pts suivis x 2 ans 1ère visite: TPP sec (autre que néo) ⇒ 0 / 107 néo TPP idiop. ⇒ 5 /153 néo (3.3%) À 2 ans: TPP sec (autre que néo) ⇒ 2 / 105 néo (1.9%) TPP idiop. ⇒ 11 / 145 néo (7.6%) ☛ p: 0.043. OR: 2.3 CI 1.0-5.2 Si récidive de TPP ( 35 sur 145 ) 6 / 35 néo (17.1%) OR: 9.8 par rapport à TPP sec. OR: 4.3 par rapport à TPP idiop. sans récidive BACKGROUND. In contrast to the established relation between overt cancer and subsequent venous thromboembolism, it is unclear whether symptomatic deep-vein thrombosis is associated with a risk of subsequent overt malignant disease. METHODS. Two hundred sixty consecutive patients with symptomatic, venographically proved deep-vein thrombosis were enrolled in a study, of whom 250 were followed during a two-year period. Among those assessed during follow-up, the incidence of subsequently detected cancer in the 105 patients with secondary venous thrombosis (i.e., thrombosis associated with a well-recognized risk factor other than cancer) was compared with the incidence of cancer in the 145 patients with idiopathic venous thrombosis. RESULTS. Routine examination at the time of diagnosis of the venous thrombosis revealed cancer in 5 of the 153 enrolled patients with idiopathic venous thrombosis (3.3 percent) and in none of the 107 enrolled patients with secondary venous thrombosis. During follow-up, overt cancer developed in 2 of the 105 patients with secondary venous thrombosis (1.9 percent) and in 11 of the 145 patients with idiopathic venous thrombosis (7.6 percent; odds ratio, 2.3; 95 percent confidence interval, 1.0 to 5.2; P = 0.043). Of the 145 patients with idiopathic venous thrombosis, 35 had confirmed recurrent thromboembolism. Overt cancer subsequently developed in 6 of the 35 (17.1 percent). The incidence of cancer in the patients with recurrent idiopathic venous thrombosis was higher than that in the patients with secondary venous thrombosis (P = 0.008; odds ratio, 9.8; 95 percent confidence interval, 1.8 to 52.2) or in the patients with idiopathic venous thrombosis that did not recur (P = 0.024; odds ratio, 4.3; 95 percent confidence interval, 1.2 to 15.3). CONCLUSIONS. There is a statistically significant and clinically important association between idiopathic venous thrombosis and the subsequent development of clinically overt cancer, especially among patients in whom venous thromboembolism recurs during follow-up

Cohorte rétrospective de 1977 à 1992 “THE RISK OF A DIAGNOSIS OF CANCER AFTER PRIMARY DEEP VENOUS THROMBOSIS OR PULMONARY EMBOLISM” Sorensen H T et al. NEJM 1998: 338: 1169-73 Cohorte rétrospective de 1977 à 1992 15,348 TPP et 11,305 EP 1,737 néo / 15,348 TPP ⇒ 11.3% OR ⇒1.3 (IC 1.21 - 1.33)

Néo: 11.3% 26,653 pts Pancréas Ovaire Foie Cerveau “THE RISK OF A DIAGNOSIS OF CANCER AFTER PRIMARY DEEP VENOUS THROMBOSIS OR PULMONARY EMBOLISM” Sorensen H T et al. NEJM 1998: 338: 1169-73 Néo: 11.3% 26,653 pts Pancréas Ovaire Foie Cerveau

Pas de différence pour l’incidence de décès dus au cancer “INCIDENCE OF CANCER AFTER PROPHYLAXIS WITH WARFARIN AGAINST RECURRENT VENOUS THROMBOEMBOLISM” Schulman S et al. NEJM 2000; 342: 1953-8 111 néo pour 854 pts traités: 13% OR: 3.4 la 1ère année (IC: 2.2 à 4.6) Suivi moyen: 8.1 années Pas de différence pour l’incidence de décès dus au cancer BackgroundThe length of time after an episodeof venous thromboembolism during which the risk ofnewly diagnosed cancer is increased is not known,and whether vitamin K antagonists have an antineoplasticeffect is controversial. MethodsIn a prospective, randomized study of theduration of oral anticoagulation (six weeks or sixmonths) after a first episode of venous thromboembolism,patients were questioned annually about anynewly diagnosed cancer. After a mean follow-up of8.1 years, we used the Swedish Cancer Registry toidentify all diagnoses of cancer and causes of deathin the study population. The observed numbers ofcases of cancer were compared with expected numbersbased on national incidence rates, and the standardizedincidence ratios were calculated. ResultsA first cancer was diagnosed in 111 of 854patients (13.0 percent) during follow-up. The standardizedincidence ratio for newly diagnosed cancerwas 3.4 (95 percent confidence interval, 2.2 to 4.6)during the first year after the thromboembolic eventand remained between 1.3 and 2.2 for the followingfive years. Cancer was diagnosed in 66 of 419 patients(15.8 percent) who were treated for six weeks withoral anticoagulants, as compared with 45 of 435 patients(10.3 percent) who were treated for six months(odds ratio, 1.6; 95 percent confidence interval, 1.1 to2.4). The difference was mainly due to the occurrenceof new urogenital cancers, of which there were 28cases in the six-week group (6.7 percent) and 12 casesin the six-month group (2.8 percent) (odds ratio, 2.5;95 percent confidence interval, 1.3 to 5.0). The differencein the incidence of cancer between the treatmentgroups became evident only after two years offollow-up, and it remained significant after adjustmentfor sex, age, and whether the thromboembolism wasidiopathic or nonidiopathic. Older age at the time ofthe venous thrombosis and an idiopathic thromboembolismwere also independent risk factors for adiagnosis of cancer. No difference in the incidence ofcancer-related deaths was detected. ConclusionsThe risk of newly diagnosed cancerafter a first episode of venous thromboembolism iselevated during at least the following two years. Subsequently,the risk seems to be lower among patientstreated with oral anticoagulants for six monthsthan among those treated for six weeks.

Incidence de nouvelle néoplasie “INCIDENCE OF CANCER AFTER PROPHYLAXIS WITH WARFARIN AGAINST RECURRENT VENOUS THROMBOEMBOLISM” Schulman S et al. NEJM 2000; 342: 1953-8 Incidence de nouvelle néoplasie BackgroundThe length of time after an episodeof venous thromboembolism during which the risk ofnewly diagnosed cancer is increased is not known,and whether vitamin K antagonists have an antineoplasticeffect is controversial. MethodsIn a prospective, randomized study of theduration of oral anticoagulation (six weeks or sixmonths) after a first episode of venous thromboembolism,patients were questioned annually about anynewly diagnosed cancer. After a mean follow-up of8.1 years, we used the Swedish Cancer Registry toidentify all diagnoses of cancer and causes of deathin the study population. The observed numbers ofcases of cancer were compared with expected numbersbased on national incidence rates, and the standardizedincidence ratios were calculated. ResultsA first cancer was diagnosed in 111 of 854patients (13.0 percent) during follow-up. The standardizedincidence ratio for newly diagnosed cancerwas 3.4 (95 percent confidence interval, 2.2 to 4.6)during the first year after the thromboembolic eventand remained between 1.3 and 2.2 for the followingfive years. Cancer was diagnosed in 66 of 419 patients(15.8 percent) who were treated for six weeks withoral anticoagulants, as compared with 45 of 435 patients(10.3 percent) who were treated for six months(odds ratio, 1.6; 95 percent confidence interval, 1.1 to2.4). The difference was mainly due to the occurrenceof new urogenital cancers, of which there were 28cases in the six-week group (6.7 percent) and 12 casesin the six-month group (2.8 percent) (odds ratio, 2.5;95 percent confidence interval, 1.3 to 5.0). The differencein the incidence of cancer between the treatmentgroups became evident only after two years offollow-up, and it remained significant after adjustmentfor sex, age, and whether the thromboembolism wasidiopathic or nonidiopathic. Older age at the time ofthe venous thrombosis and an idiopathic thromboembolismwere also independent risk factors for adiagnosis of cancer. No difference in the incidence ofcancer-related deaths was detected. ConclusionsThe risk of newly diagnosed cancerafter a first episode of venous thromboembolism iselevated during at least the following two years. Subsequently,the risk seems to be lower among patientstreated with oral anticoagulants for six monthsthan among those treated for six weeks.

“INCIDENCE OF CANCER AFTER PROPHYLAXIS WITH WARFARIN AGAINST RECURRENT VENOUS THROMBOEMBOLISM” Schulman S et al. NEJM 2000; 342: 1953-8 BackgroundThe length of time after an episodeof venous thromboembolism during which the risk ofnewly diagnosed cancer is increased is not known,and whether vitamin K antagonists have an antineoplasticeffect is controversial. MethodsIn a prospective, randomized study of theduration of oral anticoagulation (six weeks or sixmonths) after a first episode of venous thromboembolism,patients were questioned annually about anynewly diagnosed cancer. After a mean follow-up of8.1 years, we used the Swedish Cancer Registry toidentify all diagnoses of cancer and causes of deathin the study population. The observed numbers ofcases of cancer were compared with expected numbersbased on national incidence rates, and the standardizedincidence ratios were calculated. ResultsA first cancer was diagnosed in 111 of 854patients (13.0 percent) during follow-up. The standardizedincidence ratio for newly diagnosed cancerwas 3.4 (95 percent confidence interval, 2.2 to 4.6)during the first year after the thromboembolic eventand remained between 1.3 and 2.2 for the followingfive years. Cancer was diagnosed in 66 of 419 patients(15.8 percent) who were treated for six weeks withoral anticoagulants, as compared with 45 of 435 patients(10.3 percent) who were treated for six months(odds ratio, 1.6; 95 percent confidence interval, 1.1 to2.4). The difference was mainly due to the occurrenceof new urogenital cancers, of which there were 28cases in the six-week group (6.7 percent) and 12 casesin the six-month group (2.8 percent) (odds ratio, 2.5;95 percent confidence interval, 1.3 to 5.0). The differencein the incidence of cancer between the treatmentgroups became evident only after two years offollow-up, and it remained significant after adjustmentfor sex, age, and whether the thromboembolism wasidiopathic or nonidiopathic. Older age at the time ofthe venous thrombosis and an idiopathic thromboembolismwere also independent risk factors for adiagnosis of cancer. No difference in the incidence ofcancer-related deaths was detected. ConclusionsThe risk of newly diagnosed cancerafter a first episode of venous thromboembolism iselevated during at least the following two years. Subsequently,the risk seems to be lower among patientstreated with oral anticoagulants for six monthsthan among those treated for six weeks.

“INCIDENCE OF CANCER AFTER PROPHYLAXIS WITH WARFARIN AGAINST RECURRENT VENOUS THROMBOEMBOLISM” Schulman S et al. NEJM 2000; 342: 1953-8 BackgroundThe length of time after an episodeof venous thromboembolism during which the risk ofnewly diagnosed cancer is increased is not known,and whether vitamin K antagonists have an antineoplasticeffect is controversial. MethodsIn a prospective, randomized study of theduration of oral anticoagulation (six weeks or sixmonths) after a first episode of venous thromboembolism,patients were questioned annually about anynewly diagnosed cancer. After a mean follow-up of8.1 years, we used the Swedish Cancer Registry toidentify all diagnoses of cancer and causes of deathin the study population. The observed numbers ofcases of cancer were compared with expected numbersbased on national incidence rates, and the standardizedincidence ratios were calculated. ResultsA first cancer was diagnosed in 111 of 854patients (13.0 percent) during follow-up. The standardizedincidence ratio for newly diagnosed cancerwas 3.4 (95 percent confidence interval, 2.2 to 4.6)during the first year after the thromboembolic eventand remained between 1.3 and 2.2 for the followingfive years. Cancer was diagnosed in 66 of 419 patients(15.8 percent) who were treated for six weeks withoral anticoagulants, as compared with 45 of 435 patients(10.3 percent) who were treated for six months(odds ratio, 1.6; 95 percent confidence interval, 1.1 to2.4). The difference was mainly due to the occurrenceof new urogenital cancers, of which there were 28cases in the six-week group (6.7 percent) and 12 casesin the six-month group (2.8 percent) (odds ratio, 2.5;95 percent confidence interval, 1.3 to 5.0). The differencein the incidence of cancer between the treatmentgroups became evident only after two years offollow-up, and it remained significant after adjustmentfor sex, age, and whether the thromboembolism wasidiopathic or nonidiopathic. Older age at the time ofthe venous thrombosis and an idiopathic thromboembolismwere also independent risk factors for adiagnosis of cancer. No difference in the incidence ofcancer-related deaths was detected. ConclusionsThe risk of newly diagnosed cancerafter a first episode of venous thromboembolism iselevated during at least the following two years. Subsequently,the risk seems to be lower among patientstreated with oral anticoagulants for six monthsthan among those treated for six weeks.

RR de néo ↑ pour les 2 ans suivant le Dx de TPP “Excess risk of cancer in patients with primary venous thromboembolism A national, population-based cohort study” Murchison J T et al. Br J Cancer 2004; 91(1) 92-5 Cohorte écossaise 1982 - 2000 RR de néo ↑ pour les 2 ans suivant le Dx de TPP Néo ovaire + lymphome prédominants RR ↑ chez les jeunes We conducted a nationwide, retrospective cohort study assessing the risk of cancer in VTE patients diagnosed in Scotland in 1982-2000. Significantly elevated risks of cancer were sustained for 2 years after VTE diagnosis, most notably for ovarian tumours and lymphomas. Younger patients were at an increased relative risk from this association

103 pts consécutifs suivis x 12 mois 77 pts sans néo connu “Incidence and prognosis of cancer associated with bilateral venous thrombosis A prospective study of 103 patients” Bura A et al. JTH 2004; 2(3): 441-4 103 pts consécutifs suivis x 12 mois 77 pts sans néo connu 22 / 77 (26%) de ces pts ⇒ néo manifeste 40.5% si TPP idiop. 12.5% si TPP sec. 70% métastatiques Survie à 1 an 26% si néo connu 35% si néo apparu A Bura, N Cailleux, B Bienvenu, P Leger, A Bissery, H Boccalon, JN Fiessinger, H Levesque, and J Emmerich Service de M仕ecine Vasculaire-HTA and INSERM U428, H冪ital Europ仔n Georges Pompidou, Paris, France. alessandra.bura-riviere@hop.egp.ap-hop-paris.fr BACKGROUND: A strong association between bilateral deep vein thrombosis (DVT) and cancer had been found in one retrospective study. To confirm this finding, consecutive patients with an objective diagnosis of bilateral DVT were followed over 12 months. PATIENTS AND METHODS: One-hundred and three patients, hospitalized for bilateral DVT, were included in the study. Twenty-six patients (25.2%) were already known to have a cancer, 26 (25.2%) had a previous history of venous thromboembolic disease, 44 (42.7%) had a symptomatic pulmonary embolism. The patients were scheduled to be prospectively followed up at 3, 6 and 12 months as outpatients. Information on recurrence, evidence of a new overt cancer and the cause of death were recorded for all patients. RESULTS: A new cancer was diagnosed in 20 (26%) of the 77 patients without known cancer at admission. The risk of cancer was significantly more important in idiopathic thrombosis than in patients with secondary thrombosis (40.5% vs. 12.5%; odds ratio 4.8, 95% confidence interval 1.4, 18.8). Seventy percent of the cancers discovered had already spread. Age, gender, presence of pulmonary embolism, recurrence and location of the thrombosis were not statistically associated with the risk of cancer. The 1-year survival rates of patients with a previously known cancer and patients with a newly discovered cancer were, respectively, 26% and 35% (P = 0.33). CONCLUSIONS: Bilateral DVT is a significant risk indicator of malignancy. Cancer is present in 45% of patients with bilateral DVT and is associated with a poor prognosis.

Évaluation des dossiers pour néo à 1 an de suivi “In proven deep vein thrombosis, a low positive D-Dimer score is a strong negative predictor for associated malignancy” Rege K P et al TH 2004; 91: 1219-22 100 pts consécutifs Évaluation des dossiers pour néo à 1 an de suivi DD > 1000 ng/ml 22 / 66 ⇒ Néo DD < 1000 ng/ml 2 / 34 ⇒ Néo p: 0.0025 Kanchan P. Rege, Sue Jones, Jane Day, Catherine E. HoggarthDepartment of Haematology, Hinchingbrooke Hospital, Huntingdon, Cambridgeshire, UK Summary D-Dimer measurements are being increasingly used for negative prediction of deep vein thrombosis (DVT). At our institution, clinical score, D-Dimer assay, plethysmography and, if necessary, Doppler ultrasound are used to secure the diagnosis. We collected the data from 100 consecutive patients proven to have DVT. We examined their medical case notes at diagnosis for concurrent clinical conditions and one year later to look for documented evidence of malignancy. Twenty-two of the 66 patients with D-Dimers greater than 1000 ng/ml were diagnosed with a cancer compared with only 2 of the 34 patients with a presenting D-Dimer score of less than 1000 ng/ml. We propose that a D-Dimer score of less than 1000 ng/ml in proven DVT is a strong negative predictor for malignancy (p = 0.0025).

Thromboembolie veineuse et cancer Synopsis en l’absence de néo au départ OR de néo pour la TEV: 2 à 4 Risque surtout si TEV idiopathique (surtout si récurrence) Risque principalement pendant 6 à 24 premiers mois Néo: surtout… Ovaire + Uro-génital (prostate) Lymphome Cerveau Pancréas Foie Poumon et Sein: probablement évident au départ

Thromboembolie veineuse et cancer Bilan de néo à faire Anamnèse et examen physique Bilan sérologique de base et Rx poumon Examens ciblés sur éléments anormaux Examens selon les normes pour l’examen périodique annuel: Mammographie Colonoscopie ou LB ou Sang/selles Analyses spécifiques discutables: APS, CA125 etc

Thromboembolie veineuse et cancer Efficience et conscience Le statisticien: “La recherche poussée de néoplasie est peu rentable et influe peu sur le pnonostic” Le clinicien: “Le risque de néoplasie est doublé à triplé s’il y a TEV et peut atteindre 10%: le patient a le droit de choisir son devenir” Merci