Innovations: Technology & Techniques in Cardiothoracic & Vascular Surgery:
Evaluation of a Novel Cryoablation System: In vitro Testing of Heat Capacity and Freezing Temperatures
Weimar, Timo MD; Lee, Anson M. MD; Ray, Shuddhadeb MD; Schuessler, Richard B. PhD; Damiano, Ralph J. Jr MD
From the Division of Cardiothoracic Surgery, Department of Surgery, Washington University School of Medicine, Barnes-Jewish Hospital, St Louis, MO USA.
Accepted for publication December 8, 2012.
Supported, in part, by the National Heart, Lung and Blood Institute at the National Institutes of Health (grants RO1 HL032257, RO1 HL085113, and T32 HL07776).
Disclosures: Richard B. Schuessler, MD, receives research support from AtriCure, Inc, Cincinnati, OH USA. Ralph J. Damiano, MD, is a consultant for AtriCure, Inc, Cincinnati, OH USA, and Medtronic, Inc, Minneapolis, MN USA, and has received research grants from AtriCure, Inc. Timo Weimar, MD; Anson M. Lee, MD; and Shuddhadeb Ray, MD declare no conflict of interest.
Address correspondence and reprint requests to Ralph J. Damiano, Jr, MD, Division of Cardiothoracic Surgery, Washington University School of Medicine, Barnes-Jewish Hospital, Suite 3108 Queeny Tower, 1 Barnes-Jewish Hospital Plaza, St Louis, MO 63110 USA. E-mail: email@example.com.
Objective: Cryoablation has been used to ablate cardiac tissue for decades and has been shown to be able to replace incisions in the surgical treatment of atrial fibrillation. This in vitro study evaluates the performance of a novel cryoprobe and compares it with existing commercially available devices.
Methods: A new malleable 10-cm aluminum cryoprobe was compared with a rigid 3.5-cm copper linear probe using in vitro testing to evaluate performances under different thermal loads and with different tissue thicknesses. Radial dimensions of ice formation were measured in each water bath by a high-precision laser 2 minutes after the onset of cooling. Probe-surface temperatures were recorded by thermocouples. Tissue temperature was measured at depths of 4 mm and 5 mm from the probe-tissue interface. Time to reach a tissue temperature of −20°C was recorded.
Results: Ice formation increased significantly with lower water-bath temperatures (P < 0.001). Width and depth of ice formation were significantly less for the rigid linear probe (P < 0.012 and P < 0.001, respectively). There was no difference between the probes in the maximal negative temperature reached under different thermal loads or at different tissue depths. The malleable probe achieved significantly lower temperatures at the proximal compared with the distal end (−61.7°C vs −55.0°C, respectively; P < 0.001). A tissue temperature of −20°C was reached earlier at 4 mm than at 5 mm (P < 0.001) and was achieved significantly faster with the 3011 Maze Linear probe (P < 0.021).
Conclusions: The new malleable probe achieved rapid freezing to clinically relevant levels in up to 5-mm–thick tissue. Both probes maintained their performance under a wide range of thermal loads.
The mechanisms of cardiac tissue injury caused by cryothermy have been examined for more than half a century. In 1951, Taylor et al.1 reported the creation of transmural lesions in a canine model using expanding carbon dioxide to cool cardiac tissue to −60°C. Although the mode of cellular and tissue injury leading to necrotic and apoptotic cell death cascades caused by low temperature exposure was initially used exclusively in the therapy for malignancy and palliative management of pain,2 it became obvious that cardiac cells were as or more sensitive to cryoinjury than were cancerous cells.3 Cardiac tissue is destroyed by the formation of intracellular and extracellular ice crystals, which disrupt cell membranes and organelles. This had made this energy source potentially useful for the interventional treatment of cardiac arrhythmias.4–8
See accompanying editorial on page 387
As opposed to other alternative energy sources such as radiofrequency, focused ultrasound, microwave, or laser, cryoablation creates direct physical injury, cumulative sublethal cellular stress response, and molecular-based cell death by freezing.9–11 Cryodestruction occurs through immediate and delayed mechanisms of cell injury. As cardiac tissue temperature approaches −32°C, cell membranes become less fluid and transport capacity of ion pumps change, which increases action potential duration. Cardiac cellular dysfunction has been demonstrated by decreased Ca2+ sensitivity related to protein kinase A and C phosphorylation of cardiac troponin I in hypothermic cooling and rewarming experiments mimicking ischemia and reperfusion.12,13 As temperature continues to drop, intracellular pH decreases, ionic imbalance occurs, and energy levels (adenosine triphosphate) drop, leading to further membrane injury and an accumulation of free radicals. The reversibility of these cell alterations is inversely related to the degree of tissue cooling and the duration of application. Recent cryogenic techniques claim to rapidly freeze tissue down to −160°C. Ice formation is the cornerstone of tissue injury and the mode of ablation in modern devices. The onset of ice formation at the cryoprobe provides cryoadhesion to maintain and ensure tissue contact and creates an area at which heat is extracted from the tissue. As heat is removed by various cryogens such as nitrous oxide, argon, or liquid oxygen, extracellular fluid freezes at −20°C, creating a hyperosmotic environment that causes cell shrinkage and, ultimately, cell death.14 Rapid freezing to −40°C induces expansion of intracellular ice formation, which disrupts organelles and cell membranes even before osmotic imbalance occurs.10 A fast rate of cooling will increase cell death, while slowly thawing the tissue is also effective in prolonging the mechanisms of cell destruction.14
Although there are many encouraging reports of cryoablation in the surgical treatment of AF,15–17 investigations evaluating the heat capacity and the actual temperature achieved in the targeted tissue have been limited. In this study, the novel malleable nitrous oxide cryoablation system was compared with one that has been widely used for surgical cryoablation for decades. The new aluminum probe Cryo1 (AtriCure Inc, West Chester, OH USA) and the gold-plated copper 3011 Maze Linear probe (AtriCure Inc, West Chester, OH USA), formally called the Frigitronics 3011 probe, were evaluated to document their capacity to reach clinically relevant freezing temperatures over a wide range of physiological thermal loads. Neither probe has been previously examined in vitro or in vivo despite widespread clinical use.
Two cryoablation devices, the Cryo1 and the 3011 Maze Linear probe, were used in this study (Fig. 1). The Cryo1 probe is a malleable, smooth, disposable, 10-cm–long aluminum cylindrical probe with an outer diameter of 4.24 mm, an inner diameter of 2.97 mm, and a thermal conductivity coefficient of 222 W·K−1·m−1. The 3011 Maze Linear probe is a reusable device that has a rigid 3.5-cm gold-plated copper cylindrical probe with an outer diameter of 5.18 mm, an inner diameter of 3.66 mm, and a thermal conductivity coefficient of 400 W·K−1·m−1. Both probes are cooled by nitrous oxide and have an active defrost feature to free the probe from frozen tissue. Cooling and gas flow were controlled using the Cardiovascular Cryosurgical System CCS200 (AtriCure, Inc, West Chester, OH USA).
The width and the depth of ice formation on the cryoprobes in a water bath set to temperatures of 32°C, 25°C, and 10°C were measured. Each device was placed in the water bath, with the probe half submerged along its longitudinal axis, imitating the probe-tissue interface during ablation (Fig. 2). The temperatures were chosen to meet the clinical scenarios of both normothermic beating heart ablation and arrested heart ablation. The width and the depth of ice formation along the probes after 2 minutes after the onset of cooling were measured for both devices using a noncontact high-precision laser measurement system (Keyence Optical Micrometers; Keyence Corporation, Osaka, Japan). Ice formation measurements were taken at four positions along the 10-cm Cryo1 probe and at three positions along the 3.5-cm 3011 Maze Linear probe. The measurements were repeated six times for each cryoprobe.
The probe-surface temperatures were measured at four positions along the 10-cm Cryo1 probe and at three positions along the 3.5-cm 3011 Maze Linear probe by wire thermocouples (Type T; Omega Thermocouples, Stamford, CT USA) across the length of the probes, as shown in Figure 3. Temperatures were continuously recorded and plotted as a function of time during a 2-minute cooling period while the cryoprobe was placed half submerged in 32°C, 25°C, and 10°C water baths. The measurements were repeated three times for each cryoprobe.
Tissue temperatures during cryoablation were measured for both devices using fresh bovine myocardium cut in 5- and 6-mm thickness. A 4-mm spacer was used to hold the myocardial tissue off a platform. Four hypodermic thermocouples measuring 5 mm in length (Model HYP-1; Omega Thermocouples, Stamford, CT USA) extended from the platform’s base and penetrated 1 mm into the myocardial tissue. These thermocouples were positioned to lie evenly distributed and centered below the overlying cryoprobe (Fig. 4). Tissue temperatures were measured at 4 and 5 mm from the tissue-cryoprobe interface for the 5- and 6-mm–thick bovine myocardial samples, respectively. To ensure constant probe-tissue contact, a weight of 200 g was applied to the cryoprobes from above until cryoadhesion occurred. This was done to simulate the clinical use of the ablation devices. The bovine myocardial tissue samples were initially at room temperature. The spacer was removed 10 seconds after starting cryoablation when cryoadhesion was sufficient to maintain tissue-probe contact. The spacer was used to minimize thermal losses caused by tissue contact to the platform. Cryoablation was continued for 2 minutes, followed by active defrost to free the probes from the tissue. Tissue temperatures measured by each thermocouple were plotted as a function of time. For both cryoprobes, these measurements were repeated six times for each tissue thickness. On the basis of previous studies, which revealed that a temperature of −20°C was needed to cause cardiac cell death,14,18 the time needed to reach this goal temperature was identified for each ablation.
All continuous data were expressed as mean ± SD and were compared using analysis of variance testing (Systat Statistical Software, Version 13; Chicago, IL USA). Multiple comparisons were made using Fisher least significant difference test. P < 0.05 was considered significant.
The radial dimensions of ice formation on both cryoprobes increased significantly with lower water-bath temperatures (P < 0.001). The mean ± SD depth of ice formed for the Cryo1 probe was 5.41 ± 0.38 mm, 7.13 ± 0.48 mm, and 11.16 ± 0.70 mm at water-bath temperatures of 32°C, 25°C, and 10°C, respectively. There was significantly less depth of ice formed along the 3011 Maze Linear probe at water-bath temperatures of 25°C and 10°C, respectively, when compared with the Cryo1 device (P < 0.001). The mean ± SD width of ice formed along the surface of the water on either side of the Cryo1 probe was 3.03 ± 0.18 mm, 4.40 ± 0.34 mm, and 6.21 ± 0.33 mm at water-bath temperatures of 32°C, 25°C, and 10°C, respectively. The width of ice formed along the 3011 Maze Linear probe was significantly greater with 3.80 ± 0.56 mm at 32°C but was significantly less with 4.07 ± 0.65 mm and 6.93 ± 0.37 mm at water-bath temperatures of 25°C and 10°C, respectively, when compared with the Cryo1 device (P < 0.012). Overall, ice formation was homogenously distributed along the probe on both devices, with no significant differences between the proximal and distal ends (Fig. 2, Table 1).
There was no significant difference in the maximum negative temperature reached by the Cryo1 probe (−58.9°C ± 3.2°C) or the Maze Linear 3011 probe (−61.2°C ± 2.9°C, P = 0.101). The temperature of the water bath did not affect the maximum negative temperature reached by either cryoprobe (P = 0.114, Fig. 5). The maximum negative temperature measured at the Cryo1 probe surface was significantly lower at the proximal end (−61.7°C) than at the distal end (−55.0°C, P < 0.001). The same trend could be found for the 3011 Maze Linear probe, but the difference in the maximum negative temperature achieved along the probe did not reach significance for this device (proximal, −63.3°C; distal, −58.3°C; P = 0.058).
The mean ± SD tissue temperatures measured from the thermocouples along each probe’s length after 2 minutes of ablation at a distance of 4 mm from the probe-tissue interface were −34.3°C ± 6.3°C for the Cryo1 probe and −36.6°C ± 6.6°C for the 3011 Maze Linear probe (Fig. 6). There was a trend for a lower temperature measured at a distance of 5 mm from the probe-tissue interface, with a mean ± SD of −28.9°C ± 3.7°C for the Cryo1 probe and −31.0°C ± 8.6°C for the Maze Linear 3011 device (P = 0.071). The maximum negative tissue temperature achieved was not significantly affected by the device used (P = 0.439) or the position of measurement (P = 0.750). The measured tissue temperatures beneath the distal and the proximal end of the devices did not show significant differences for either device (P = 0.205). For the Cryo1 probe, the time of cryoablation necessary to attain a tissue temperature of −20°C was 60.5 ± 25.8 seconds at a 4-mm distance and 86.0 ± 9.3 seconds at a 5-mm distance from the tissue-probe interface. For the Maze Linear 3011 device, 39.4 ± 15.9 seconds at a 4-mm distance and 73.5 ± 21.8 seconds at a 5-mm distance were needed to achieve a temperature of −20°C. Overall, the goal temperature of −20°C was reached by both devices, significantly faster at a 4-mm than a 5-mm distance (P < 0.001), but was achieved significantly faster by the 3011 Maze Linear probe than by the Cryo1 probe (P < 0.021).
The mechanisms of cell injury during cryoablation are complex. Intracellular and extracellular ice formation causes a hyperosmotic environment, eventually leading to either cell death or disruption of organelles and cell membranes. Earlier studies have suggested a tissue temperature of −20°C to be sufficient to induce necroses.10,18 To accomplish low temperatures, modern systems allow compressed gas to expand into a lower-pressure environment within the probe, called the Joule-Thomson effect. Newton’s law of cooling can evaluate the rate at which heat (Q˙) is removed from a system by the working fluid as a function of change in enthalpy (Δh) and the mass flow rate (m˙):
Equation (Uncited)Image Tools
. The chosen coolant can impact the level of heat removal in the analysis of enthalpy, which is a function of the specific heat (Cp) and change in temperature (ΔT). Two widely used refrigerants in cardiac procedures are monatomic argon and polyatomic nitrous oxide. A measure of the heat that is required to raise the substance’s temperature of a unit quantity by a unit degree is the specific heat. Nitrous oxide as a polyatomic gas has multiple free vibrational modes resulting in a high specific heat of 0.890 kJ/(kg·C) and is therefore capable of absorbing more heat with a smaller rise in temperature than are other refrigerants. Thus, the amount of heat energy that is removed from a system is contingent on the specific heat of the cooling fluid, the final temperature of the cooling fluid after the throttling process, and the mass flow rate of the fluid through the system. The optimal freeze-thaw cycle of an ablation device should consist of a fast cooling rate and a slow thawing rate to maximize tissue injury.14,19 In this study, our results revealed the capability of the nitrous oxide system to rapidly cool down both probes tested to a goal temperature of −20°C, which is felt to ensure induction of cell necrosis.
In addition to cooling capability, probe material and surface contact with the targeted tissue impact heat conduction. The probe design strongly influences the thermal resistance (Rth) of a system when considering the equation
Equation (Uncited)Image Tools
where t is the probe-wall thickness, A is the contact surface area, and k represents the thermal conductivity of the probe material. The thermal conduction coefficient (k) is the prediction of the rate of heat transfer or heat loss, respectively, through a unit thickness of material. Copper, which is used as the probe material in the 3011 Maze Linear probe device, has a thermal conduction coefficient of 400 W·K−1·m−1 and was expected to extract heat faster from the targeted tissue than the aluminum probe of the Cryo1 device with a thermal conduction coefficient of only 222 W·K−1·m−1.
In our study, the maximal negative temperature achieved in the tissue was similar for both devices, but the 3011 Maze Linear probe reached a goal temperature of −20°C significantly faster than did the Cryo1 probe according to its two times higher thermal conduction coefficient. However, our results did not show a significant difference in negative tissue temperature reached at a depth of 4 or 5 mm over the examined length of about 3.5 cm for this device, and the maximal negative tissue temperatures reached within 2 minutes of ablation were well below the critical −20°C mark. This indicates a similar freezing performance with an increase in probe malleability and lesion length of the new 10-cm–long probe. Importantly, the ice formation measured in the water-bath testing was homogenous along the entire Cryo1 probe. This indicated homogeneous cooling, which can be an issue for long, flexible probes. There were no significant differences in the probe-surface temperatures between the two probes.
The wall thickness of the human cardiac atria differs depending on region, coexisting cardiac disease, or age,20–22 but is an important predictor for failure in creating transmural lesions.23 In this study, temperature was measured at tissue depths of 4 and 5 mm, chosen because these approximated the average wall thickness of most human atria.22 The maximum negative temperature after 2 minutes of ablation was significantly lower at a 4-mm than a 5-mm distance from the probe-tissue–contact interface for both devices. As the rate of cooling and the maximum negative temperature decrease with increasing distance from the probe-tissue interface, extracellular ice formation results in a zone of incomplete and reversible tissue damage.24 This confirms the findings of earlier animal studies revealing differences in the performance of cryoablation depending on the targeted atrial wall structures and thicknesses.15,23,25 Our findings also suggest that in thick atrial tissue greater than 5 mm, ablation times of longer than 2 minutes are warranted. There are limitations to this study. It used an in vitro model, which allowed for accurate measurement of probe temperature during ice formation, but did not fully simulate living tissue. However, this study was designed to assess the relative performance of the two probes in a well-controlled environment. Another limitation is that the laboratory setup may not fully replicate clinical use.
It is extremely important to have uniform probe-tissue contact. Ice formation caused by fluid or vapor trapped in the valleys of a nonuniform surface is a problem because ice has a low thermal conductivity of 3 W·K−1·m−1, and, if introduced to the system, will increase thermal resistance and become an insulator. Both devices tested were loaded with a weight of 200 g after establishing sufficient tissue-probe contact, which was felt to mimic realistically the pressure applied by the surgeon in the clinical application. The setting used in this study represents a nearly optimal probe-surface contact that may be difficult to reproduce in surgery.
In conclusion, the nitrous oxide system with the new malleable aluminum probe of the Cryo1 device and the rigid copper probe of the 3011 Maze Linear probe achieved rapid freezing to clinically relevant levels in up to 5-mm–thick myocardial tissue in this in vitro model. The performance of the two probes was excellent even when burdened with different thermal loads. Freezing capacity dropped, as expected, with increasing tissue thickness. However, the critical freezing temperature to induce tissue necrosis was reached within 2 minutes.
1. Taylor CB, Davis CB Jr, Vawter GF, Hass GM. Controlled myocardial injury produced by a hypothermal method. Circulation
. 1951; 3: 239–253.
2. Steinbach JP, Weissenberger J, Aguzzi A. Distinct phases of cryogenic tissue damage in the cerebral cortex of wild-type and c-fos deficient mice. Neuropathol Appl Neurobiol
. 1999; 25: 468–480.
3. Gage AA, Baust JM, Baust JG. Experimental cryosurgery investigations in vivo. Cryobiology
. 2009; 59: 229–243.
4. Khargi K, Keyhan-Falsafi A, Hutten BA, Ramanna H, Lemke B, Deneke T. Surgical treatment of atrial fibrillation: a systematic review. Herzschrittmacherther Elektrophysiol
. 2007; 18: 68–76.
5. Lall SC, Damiano RJ Jr. Surgical ablation devices for atrial fibrillation. J Interv Card Electrophysiol
. 2007; 20: 73–82.
6. Holman WL, Ikeshita M, Douglas JM Jr, Smith PK, Cox JL. Cardiac cryosurgery: effects of myocardial temperature on cryolesion size. Surgery
. 1983; 93: 268–272.
7. Ahmed H, Neuzil P, Skoda J, et al.. The permanency of pulmonary vein isolation using a balloon cryoablation catheter. J Cardiovasc Electrophysiol
. 2010; 21: 731–737.
8. Tang M, Kriatselis C, Nedios S, et al.. A novel cryoballoon technique for mapping and isolating pulmonary veins: a feasibility and efficacy study. J Cardiovasc Electrophysiol
. 2010; 21: 626–631.
9. Hass GM, Taylor CB. Quantitative hypothermal method for production of local injury to tissue. Proc Inst Med Chic
. 1947; 16: 424.
10. Gage AA, Baust J. Mechanisms of tissue injury in cryosurgery. Cryobiology
. 1998; 37: 171–186.
11. Schacht V, Becker K, Szeimies RM, Abels C. Apoptosis and leucocyte-endothelium interactions contribute to the delayed effects of cryotherapy on tumours in vivo. Arch Dermatol Res
. 2002; 294: 341–348.
12. Han YS, Tveita T, Prakash YS, Sieck GC. Mechanisms underlying hypothermia-induced cardiac contractile dysfunction. Am J Physiol Heart Circ Physiol
. 2010; 298: H890–H897.
13. Layland J, Solaro RJ, Shah AM. Regulation of cardiac contractile function by troponin I phosphorylation. Cardiovasc Res
. 2005; 66: 12–21.
14. Mazur P. Cryobiology: the freezing of biological systems. Science
. 1970; 168: 939–949.
15. Milla F, Skubas N, Briggs WM, et al.. Epicardial beating heart cryoablation using a novel argon-based cryoclamp and linear probe. J Thorac Cardiovasc Surg
. 2006; 131: 403–411.
16. Doll N, Kiaii BB, Fabricius AM, et al.. Intraoperative left atrial ablation (for atrial fibrillation) using a new argon cryocatheter: early clinical experience. Ann Thorac Surg
. 2003; 76: 1711–1715.
17. Lustgarten DL, Keane D, Ruskin J. Cryothermal ablation: mechanism of tissue injury and current experience in the treatment of tachyarrhythmias. Prog Cardiovasc Dis
. 1999; 41: 481–498.
18. Cooper IS. Cryobiology as viewed by the surgeon. Cryobiology
. 1964; 51: 44–51.
19. Gage AA, Baust JG. Cryosurgery for tumors—a clinical overview. Technol Cancer Res Treat
. 2004; 3: 187–199.
20. Saremi F, Channual S, Krishnan S, Gurudevan SV, Narula J, Abolhoda A. Bachmann Bundle and its arterial supply: imaging with multidetector CT—implications for interatrial conduction abnormalities and arrhythmias. Radiology
. 2008; 248: 447–457.
21. Ren JF, Marchlinski FE, Callans DJ, Zado ES. Echocardiographic lesion characteristics associated with successful ablation of inappropriate sinus tachycardia. J Cardiovasc Electrophysiol
. 2001; 12: 814–818.
22. Pan NH, Tsao HM, Chang NC, Chen YJ, Chen SA. Aging dilates atrium and pulmonary veins: implications for the genesis of atrial fibrillation. Chest
. 2008; 133: 190–196.
23. Doll N, Kornherr P, Aupperle H, et al.. Epicardial treatment of atrial fibrillation using cryoablation in an acute off-pump sheep model. Thorac Cardiovasc Surg
. 2003; 51: 267–273.
24. Clarke DM, Robilotto AT, Rhee E, et al.. Cryoablation of renal cancer: variables involved in freezing-induced cell death. Technol Cancer Res Treat
. 2007; 6: 69–79.
25. Guiraudon GM, Jones DL, Skanes AC, et al.. En bloc exclusion of the pulmonary vein region in the pig using off pump, beating, intra-cardiac surgery: a pilot study of minimally invasive surgery for atrial fibrillation. Ann Thorac Surg
. 2005; 80: 1417–1423.
This is the first of two articles in this issue of Innovations comparing the performance of a new, flexible, disposable cryoprobe with a previously available, rigid, reusable probe. Using nitrous oxide as the refrigerant for both probes, this in vitro study demonstrates that both probes reliably achieve a cell killing temperature of minus 20 degrees Centrigrade with 2 minute freezes of 5mm sections of bovine myocardium. However, the margin of efficacy is relatively narrow in that the flexible probe requires 86 seconds to reach this critical temperature. Given that human atria, particularly in the mitral isthmus/atrioventricular groove area, can be substantially thicker than 5 mm, the application time of the probe may require significantly longer times than the 2 minutes recommended by the manufacturer. In addition, this supports the fact that endocardial and epicardial applications that “mirror” each other across the mitral isthmus are required to reliably achieve a uniform, transmural lesion.
Future studies should examine the performance characteristics of the flexible nitrous oxide probe with the flexible argon probe. Another practical point in the article is the importance of avoiding “pleats” when applying the probe to the atrium. By virtue of its low thermal conductivity, the ice produced in the pleats will act as an insulator and possibly prevent formation of a uniform, transmural lesion. All in all, this paper is an important contribution to our understanding of cryothermia as a power source for the surgical treatment of cardiac arrythmias.
Harold G. Roberts, MD, is the guest editor.
Ablative therapy; Arrhythmia therapy; Atrial fibrillation
Copyright © 2012 by the International Society for Minimally Invasive Cardiothoracic Surgery. Unauthorized reproduction of this article is prohibited.
What does "Remember me" mean?
By checking this box, you'll stay logged in until you logout. You'll get easier access to your articles, collections,
media, and all your other content, even if you close your browser or shut down your
To protect your most sensitive data and activities (like changing your password),
we'll ask you to re-enter your password when you access these services.
What if I'm on a computer that I share with others?
If you're using a public computer or you share this computer with others, we recommend
that you uncheck the "Remember me" box.
Highlight selected keywords in the article text.
Data is temporarily unavailable. Please try again soon.