[Incidence and distribution of 1026 lower limb venous thrombi diagnosed by ultrasonography in 3263 patients with cancer]. / Incidence et distribution de 1026 thromboses veineuses des membres inférieurs diagnostiquées par écho-Doppler chez 3263 patients atteints de cancer.

UNLABELLED: The study objective was to evaluate the incidence, characteristics and distribution of lower limb deep venous thrombi diagnosed by ultrasound among known cancer patients, in order to determine additional criteria that could be used to help select patients with unprovoked deep venous thrombosis caused by occult cancer. PATIENTS AND METHODS: From January 2001 to May 2012, standardized report forms (n=38,424) were entered into a database. A total of 3263 report forms corresponding to cancer patients with suspected venous thrombo-embolic disease were selected from which 1026 cases of deep venous thrombosis were diagnosed by ultrasonography. Baseline characteristics, incidence, and anatomic distribution of venous thrombi were analyzed. RESULTS: For 3263 patients with cancer, the venous thrombosis incidence was: total 31.4% (n=1026), proximal 14.5% (n=472), bilateral 8.5% (n=278) and multiple venous sites 4.6% (n=149). The rate of clinical suspicion of pulmonary embolism was 49.9% (n=1628). For 1026 patients with thrombosis, proximal thrombi were nearly as frequent as distal thrombi, with 17.6% (n=181) iliocaval thrombi. Gastrocnemial, popliteal and femoral veins were almost equally concerned by thrombosis with respective rates of 28.7% (n=278), 27.1% (n=294) and 25.6% (n=263). Superficial veins were concerned in 23.5% (n=241). Partial or floating clots occurred frequently in 4 localizations: common femoral, external iliac, femoral and popliteal veins. CONCLUSION: Proximal, multiple, partial, mobile thrombi, and such unusual locations as gastrocnemial or superficial thromboses, are potentially indicators for selecting patients that may benefit from a cancer check-up because their venous thrombosis could be due to cancer.

Determination of refractive-index profiles by a combination of visible and infrared ellipsometry measurements.

Model inhomogeneous silicon oxynitride films were produced by ion-beam sputtering and characterized by ellipsometry in the visible and infrared ranges. These films exhibit strong intentional gradients of the refractive index that cannot be considered linear. A discrete description of the index profile with a few layers or a continuous description with a polynomial are examined; regressions by the use of measurements in the visible only or in the total spectral range (visible and infrared) are performed. Acquisition of data in an extended range is found to be a guarantee of the reliability of the calculated index profiles.