Living-Related Donor Lobectomy for Bilateral Lobar Transplantation in Patients With Cvstic Fibrosis J
Robbin G. Cohen, MD, Mark L. Barr, MD, Felicia A. Schenkel, RN, Tom R. DeMeester, MD, Winfield J. Wells, MD, and Vaughn A. Starnes, MD USC Cardiothoracic Center, Department of Surgery, University of Southern California School of Medicine, Los Angeles, California
Donor lobectomy has been performed in 14 patients enabling 7 recipients with cystic fibrosis to undergo bilateral living-related lobar pulmonary transplantation. Donors included 11 parents, 2 brothers, and 1 uncle. Donor mean age was 43 years (range 24 to 55 years). Their mean height and weight was 170 cm (range, 169 to 180 cm) and 72.4 kg (range, 55 to 90 kg), respectively, compared with 161 cm (range, 140 to 175 cm) and 42.4 kg (range, 27 to 55 kg), respectively, in the recipient group. Donor pulmonary evaluation consisted of a history and physical examination, chest roentgenogram and computed tomographic scan, spirometry with arterial blood gas measurement, echocardiography, and perfusion scanning. From each pair of donors, one was selected for right lower lobectomy and the other for left lower lobec-
tomy. Standard lobectomy techniques were modified to facilitate implantation and optimize preservation of the donor lobes. On the right side, the middle lobe was removed and discarded in the first three donors to provide an adequate cuff of pulmonary artery and bronchus for implantation. With increased experience, this has proved not to be necessary. There have been no deaths and no long-term complications in the donor group. Prolonged postoperative air leaks occurred in the 3 patients who underwent right lower and middle lobectomies. All donors have been able to resume their previous lifestyles. Living-related donor lobectomy provides an alternative to cadaveric organs in select patients in need of pulmonary transplantation. (Ann Thorac Surg 1994;57:1423-8)
L
ment of the lobes being considered. Serologic tests included those for hepatitis (A, B, and C), human immunodeficiency virus, and cytomegalovirus. A prospective crossmatch was performed between the recipient and potential donors. In each case, potential donors were interviewed by an independent committee and a psychiatrist to safeguard against coercion and to ensure donor comprehension of the procedure. Criteria for donor acceptance are as follows:
obar pulmonary transplantation at our institution has expanded from implantation of a unilateral lobe in the neonatallpediatric patient to bilateral living-related lobar transplantation in selected patients with cystic fibrosis [l]. Donor lobectomies for living-related lobar transplantation require a thorough evaluation of potential donors, as well as modification of standard lobectomy and pulmonary preservation techniques. This report summarizes the donor evaluation, surgical technique, and results of our experience with living-related donor lobectomies for pulmonary transplantation.
Material and Methods Since January 1993, 14 living-related donor lobectomies have been performed for implantation into 7 recipients with cystic fibrosis. Eleven donors were the parents of recipients, 2 were brothers, and 1 was an uncle. After determination of ABO blood group compatibility with the potential recipient, the donor evaluation consisted of a room air arterial blood gas measurement, spirometry, echocardiography, electrocardiography, ventilatiodperfusion scan, chest film, and computed tomography of the chest to exclude pathology and allow volumetric assessPresented at the Thirtieth Annual Meeting of The Society of Thoracic Surgeons, New Orleans, LA, Jan 31-Feb 2, 1994. Address reprint requests to Dr Cohen, USC Cardiothoracic Center, Department of Surgery, University of Southern California School of Medicine, 1510 San Pablo St, Suite 415, Los Angeles, CA 90033-4612.
0 1994 by The Society of Thoracic Surgeons
Age 5 55 years No significant past medical history No recent viral infections Normal echocardiogram Normal electrocardiogram Normal chest roentgenogram Oxygen tension > 80 mm Hg on room air Forced expiratory volume in 1 second and forced vital capacity > 85% predicted No significant pulmonary pathology on computed tomography (completely normal on donor side) No previous thoracic operation on donor side After a suitable donor pair was selected, one donor was selected for removal of the right lower lobe, and the other for removal of the left lower lobe. If an otherwise acceptable donor had a history of prior unilateral thoracic operation, trauma, or infection, the contralateral side was used for donation. Because the right lower lobe is smaller 0003-4975/94/$7.00
1424
COHEN ET AL LIVING-RELATED DONOR LOBECTOMY
supermr segmental
artery 8 bronchus
Fig 1. Dissection and division of the pulmonary artery for donor right lower lobectomy.
than the left lower lobe, the larger donor was selected for donation of the right lower lobe. If both donors were the same size, the side of donation was assigned arbitrarily.
Surgical Technique Three operating rooms and surgical teams are used for living-related bilateral lobar transplants. Epidural catheters for postoperative analgesia are inserted before anesthetic induction. After induction of general anesthesia, fiberoptic bronchoscopy is performed to exclude evidence of infection or inflammation, and unexpected alterations in bronchial anatomy. The endotracheal tube is replaced with a double-lumen endotracheal tube, and the patient is placed in the appropriate lateral decubitus position. The lung on the donor side is deflated. The donor lobectomy is performed through a posterolateral thoracotomy incision to maximize exposure of the hilum without manipulation or clamping of the graft. As opposed to lobectomies performed for tumor or infection, the donor lobe must be excised with an adequate cuff of pulmonary artery, pulmonary vein, and donor bronchus to allow for implantation into the recipient. In general, all dissection is camed out on the side of the remaining lobe to minimize air leaks in the recipient. DONOR RIGHT LOWER LOBECTOMY. After the chest is opened, the lung is carefully inspected to exclude pathology that might have been missed with the preoperative evaluation. The inferior pulmonary ligament is incised with the cautery as in a standard lobectomy. The mediastinal pleura is dissected anteriorly as far superiorly as the superior pulmonary vein, and posteriorly to the inferior aspect of the takeoff of the right upper lobe bronchus. Dissection in the fissure is carried out to isolate the pulmonary artery to the right lower lobe, and to define the anatomy of the pulmonary arteries to the right middle lobe. The most desirable anatomy is when there is ade-
Ann Thorac Surg 1994;571423-8
quate distance between the takeoff of the middle lobe artery and the pulmonary artery to the superior segment of the right lower lobe. This allows placement of a vascular clamp below the middle lobe artery, leaving a sufficient vascular cuff for the pulmonary arterial anastomosis at implantation (Fig 1). After confirming that the venous drainage from the right middle lobe does not enter the inferior pulmonary vein, the surgeon incises the pericardium surrounding the inferior pulmonary vein. This allows a vascular clamp to be placed on the left atrium so that an adequate pulmonary venous cuff remains on the donor lobe (Fig 2). Once the vascular dissections are complete, the fissures are stapled using a 75-mm GIA stapler (US Surgical Corp, Nonvalk, CT) and raw areas of pulmonary parenchyma are cauterized. After the lobar dissection is completed, the lung is reinflated by the anesthesiologist. Ten thousand units of heparin and 500 mg of methylprednisolone are administered intravenously, and the lung is ventilated for 5 to 10 minutes. The lung then is deflated on the donor side and a vascular clamp is placed on the pulmonary artery above the planned point of transection. A second vascular clamp then is placed on the left atrial side of the inferior pulmonary vein (see Figs 1, 2). Placing the clamps in this sequence avoids vascular congestion of the pulmonary graft. The pulmonary artery then is transected with an adequate vascular cuff for the anastomosis to the recipient without compromising the artery to the superior segment of the lower lobe. An adequate length of pulmonary artery must remain to repair it without compromising remaining pulmonary arterial branches. The inferior pulmonary vein then is transected with a small cuff of left atrium. Attention then is focused on the bronchus to the right lower lobe, which is now exposed underneath the transected pulmonary artery. As little dissection as possible is performed around the bronchus to preserve the blood supply of both the bronchus on the donor lobe and that of the remaining lung. The right middle lobe bronchus is identified, and a no. 15 blade is used to transect
right
lung I ,
”
’‘
pericardium
Fig 2. Dissection of the right inferior pulmonary vein so that a vascular clamp can be placed on the intrapericardial left atrium.
Ann Thorac Surg 1994;571423-8
COHEN ET AL LIVING-RELATED DONOR LOBECTOMY
1425
afiery to segment of superior lower lobe
Fig 3. Dissection and division of the bronchus to the right lower lobe.
the bronchus to the lower lobe. The incision starts in the bronchus intermedius above the bronchus to the superior segment of the right lower lobe, and moves obliquely to just below the takeoff of the right middle lobe bronchus (Fig 3). The lobe then is wrapped in a cold, moist sponge and taken to a separate sterile table for preservation. The donor’s pulmonary artery is repaired with running 6-0 polypropylene. The left atrium is closed with 4-0 polypropylene, and the bronchus is closed with interrupted simple sutures of 5-0 polypropylene. The chest then is closed in the traditional manner. The chest is opened and the inferior pulmonary ligament incised as for the right side. The dissection of the pulmonary artery to the lower lobe is performed so that a vascular clamp can be placed proximal to the pulmonary artery to the superior segment of the lower lobe (Fig 4). If a lingular artery takes off too far distal to the artery to the superior segment of the lower lobe and is of small size, it is ligated and divided. The pericardium is opened circumferentially around the inferior pulmonary vein. Fissures are completed with the GIA stapler. At the completion of the dissection, the lung is inflated, and heparin and methylprednisolone are administered as for the right side. The left lung then is deflated, and the pulmonary artery and vein are clamped and transected. The exposed bronchus to the left lower lobe is followed upward until the lingular bronchus is identified. The main bronchus is transected tangentially starting at the base of the upper lobe bronchus and ending 2 cm proximally, approximately 3 mm superior to the bronchus to the superior segment of the left lower lobe (Fig 5). The left lower lobe then is taken to a separate table for preservation and storage, and the pulmonary vessels and bronchus are repaired in a fashion similar to that for right lower lobectomy.
Fig 4. Dissection and division of the pulmonary artery for donor left lower lobectomy.
Pulmonary Preservation An intravenous drip of prostaglandin is initiated at the begnning of the donor operation to maximally dilate the pulmonary vasculature. The drip is titrated to a systolic blood pressure of 110 mm Hg. Once the donor lobes are excised, they are taken to a separate sterile field, where they are quickly immersed in cold crystalloid. The bronchus is carefully intubated with a small endotracheal tube,
DONOR LEFT LOWER LOBECTOMY.
Fig 5. Dissection and division of the bronchus to the left lower lobe.
1426
COHEN ET AL LIVING-RELATED DONOR LOBECTOMY
Ann Thorac Surg
1994;571423-8
Table 1 . Characteristics of Donors and Recipients for LivingRelated Lobar Transplantation Age (y)
Recipients Donors Height (cm)/ Age Height (cm)/ Lobe weight (kg) Relationship (y) weight (kg) Donated
21
157140
Mother
Father Mother Father Mother Father
13
140/27
22
167/46
21
167145
Mother Father
22
175148
Brother
23
165155
24
157136
Brother Father Mother Father
Uncle LLL = left lower lobe; lobe.
49 56 38 39 48 47 40 44 28 24 49 47 52 41
RLL = right lower lobe;
165/67 180190 159156 169173 178161 175/75 162/55 180188 175166 175179 185187 157/59 165/80 167179
RLLIRML LLL RLLIRML LLL RLLIRML LLL LLL RLL RLL LLL RLL LLL LLL RLL
RML = right middle
and the lobe is gently ventilated with 100% oxygen. The pulmonary artery is intubated and the lobe flushed with at least 1 L of cold modified Euro-Collins solution so that the pulmonary venous effluent becomes clear, and the pulmonary parenchyma is homogeneously white. Care is taken not to allow either the crystalloid bath or the preservation solution to enter the bronchus. The donor lobe is then placed in cold storage and transported to the recipient’s operating room for implantation.
Results The characteristics of the donors and recipients as well as lobes transplanted are listed in Table 1. There were 7 left lower lobectomies, 3 right lower and middle lobectomies, and 4 right lower lobectomies. All donor lobes have functioned adequately in the recipients, and there have been no recipient deaths. The mean postoperative hospital stay of the donors was 6.3 days (range, 4 to 11 days) for right lower lobectomies, 24 days (range, 18 to 28 days) for right middle and lower lobectomies, and 10.3 days (range, 5 to 14 days) for left lower lobectomies. All patients were discharged from the hospital within 48 hours of having their chest tubes removed. Prolonged hospital stays were always the result of prolonged air leaks or persistent chest tube drainage. Prolonged postoperative air leaks occurred in all 3 patients who underwent right lower and middle lobectomies (18 to 21 days), 1 patient who underwent right lower lobectomy (11 days), and 1 patient who underwent left lower lobectomy (14 days). All resolved with chest tube suction. The first donor who underwent a right middle and lower lobectomy required drainage of a suspected empyema (subsequently culture negative) on the 25th postoperative day. Transient atrial fibrillation developed
in 1patient on postoperative day 3. He converted to sinus rhythm with medical therapy and was discharged from the hospital on postoperative day 5. There have been no deaths and no long-term complications in the donor group. Postoperative spirometry at 1 to 6 months (mean follow-up, 3 months) in the first 8 donors is listed in Table 2. All donors have been able to resume their previous lifestyles.
Comment As recipient lists for patients in need of lung transplantation continue to grow, donor availability remains static [2]. The result is that many patients with end-stage pulmonary disease due to cystic fibrosis will die while waiting for donor lungs. Experimental evidence, as well as our early clinical experience, suggests that lobar pulmonary transplantation is technically feasible and can provide adequate pulmonary function in selected patients [l, 31. To date, we have performed 8 bilateral lobar pulmonary transplantations, of which 7 were for pulmonary failure due to cystic fibrosis. Although results are preliminary, all 7 patients with cystic fibrosis have survived. The eighth recipient, a 9-year-old girl with obliterative bronchiolitis, died of pneumonia secondary to adenovirus on postoperative day 22. Donor evaluation and selection for living-related transplantation involves careful assessment of both physiologic and psychosocial factors. Potential donors must have normal pulmonary function and be without other pulmonary pathology as demonstrated by spirometry and roentgenographic examinations. The minimal required recipient to donor ratio of height and weight to provide adequate pulmonary function in the recipient has yet to be determined. Methods for volumetric assessment of donor lobes such as spiral computed tomographic scanning currently are under investigation. Donors must otherwise be in excellent health without factors that might increase the risk of the donor operation. All potential donors are thoroughly evaluated as to their emotional stability, comprehension of the donor procedure, and motivation for donating to safeguard against coercion.
Table 2. Decrease in FVC and FEV, After Donor Lobectomy According to Lobes Removed Lobectomy LLL (n = 4) RLL (n = 1) RLL/RML (n = 3)
Decrease in FVC (L)
Decrease in FEV,
% Decrease
0.9 f 0.5 18% 4 10% 0.7 f 0.3
20% f 10%
1.16
% Decrease
in FVC
17%
1.67 f 0.2 40% 2 2%
(L)
0.9
in FEV,
17%
1.0 t 0.13 33% f 3.2%
FEV, = forced expiratory volume in 1 second; FVC = forced vital LLL = left lower lobe; RLL = right lower lobe; Rh4L = capacity; right middle lobe.
COHEN ET AL LIVING-RELATED DONOR LOBECTOMY
Ann Thorac Surg
1994;571423-8
Originally, only parents of recipients were seriously considered for organ donation. We have since expanded our donor criteria to include 2 brothers of 1 recipient, and an uncle of another, when it became clear that the recipients’ parents were not acceptable donors. Although we have had occasion to transplant each of the pulmonary lobes, the right lower lobe appears to be most suitable for the recipient’s right side, and the left lower lobe most suitable for the recipient’s left chest when bilateral lobar transplantation is to be performed. Each of these lobes can be removed with an adequate bronchus, pulmonary artery, and pulmonary vein to allow for functional anastomoses to the recipient main pulmonary artery, superior pulmonary vein, and main bronchus. The details of the recipient operation have been described previously [ 11. Several general principles apply to donor lobe operations that differ from those of pulmonary resections for neoplasm or infection. As opposed to smaller, musclesparing incisions, which are now commonly used for standard pulmonary resections, a posterolateral thoracotomy is used to facilitate exposure of the pulmonary arteries and intrapericardial pulmonary vein. Care is taken to avoid excessive manipulation of the donor lobe including the use of lung clamps for traction. Dissection in the pulmonary fissures is carried out on the side of the remaining lung to avoid air leaks in the recipient. Fissures are closed with surgical staplers and cautery to further avoid air leaks in both donors and recipients. Due to the proximity of the pulmonary artery and bronchus of the right middle lobe to those of the superior segment of the right lower lobe, the right middle lobe was removed and discarded in the first 3 cases to provide adequate cuffs for implantation. Prolonged air leak in these patients occurred and was attributed to a combination of the space problem created by the extended pulmonary resections and the fact that most of the anatomic dissection was carried out on the side of the remaining lung to minimize air leaks in the recipients. Further experience with both the donor and recipient operations has shown that bronchial and pulmonary arterial cuffs of only 2 to 3 mm beyond the branches are sufficient for implantation. Since then we have successfully performed donor right lower lobectomies without removing the middle lobe in the last 4 cases by accepting shorter cuffs on the pulmonary artery and bronchus of the donor right lower lobe. These donors have had routine postoperative courses, and 3 of 4 patients have been discharged from the hospital within 6 days of their operations. There have been no bronchial or vascular anastomotic complications in the recipients. For anatomic reasons, donor left lower lobectomy tends to be a simpler procedure than donor right lower lobectomy. Occasionally, a lingular artery is ligated if its takeoff is significantly distal to the artery to the superior segment of the lower lobe. On one occasion, two lingular arteries were removed as a Carrel patch and reimplanted into the remaining pulmonary artery.
1427
Long-term morbidity after donor lobectomy has been minimal. All patients have fully recovered and have returned to their previous lifestyles. The results of postoperative spirometry have been consistent with those in patients undergoing lobectomies for other reasons [4]. Donors who underwent right lower and middle lobectomies had a significantly greater drop in postoperative spirometry than those undergoing right or left lower lobectomy alone. This further emphasizes the need to protect the right middle lobe when performing donor right lower lobectomy. Other, more precise methods for determining the long-term effects of donor lobectomy are needed. We have recently begun using preoperative and postoperative exercise testing in donors to identify changes in exercise capacity as a result of donor lobectomy. Pulmonary preservation techniques commonly used in the transplantation of cadaveric lungs require modification for living-related transplantation. Neither topical cooling nor the infusion of high-potassium preservation solutions is feasible while the donor lobes remain in situ. We have developed a technique whereby the livingrelated donor lobe is quickly moved to a separate sterile table in the operating room, where it is immediately immersed in cold topical crystalloid solution. The lobar bronchus is gently intubated to ventilate the lobe. The pulmonary artery is similarly cannulated, and the lobe is flushed with the preservation solution until the pulmonary venous effluent is clear. Usually, 1 to 2 L of pulmonoplegia is required. Care is taken to see that all of the lobar parenchyma is ventilated and that the parenchyma has turned from pink to homogeneous white. Because the bronchial and pulmonary arterial cuffs are short, it often is necessary to selectively intubate segmental bronchi or arteries to assure adequate pulmonary preservation. We are careful to prevent both the pulmonoplegia and topical cold solution from entering the airways of the lobar graft. Living-related pulmonary lobar transplantation remains investigational. Due to the critical donor shortage, we have been encouraged with our preliminary results. Donor left and right lower lobectomies for living-related bilateral lobar transplants provide a source of donor lungs for select patients in need of bilateral lung transplantation. They can be performed safely with acceptable results.
References 1. Starnes VA, Barr ML, Cohen RG. Lobar transplantation: indications, technique, and outcome. J Thorac Cardiovasc Surg (in press). 2. Evans RW. The actual and potential supply of organ donors in the United States. Clin Transplants 1990;32941. 3. Backer CL, Ohtake S, Zales VR, et al. Living-relatedlobar lung transplantation in beagle puppies. J Pediatr Surg 1991;26: 429-32. 4. Ali MK, Mountain CF, Ewer MS, et al. Predicting loss of pulmonary function after pulmonary resection for brochogenic carcinoma. Chest 1980;7733742.
1428
COHEN ET AL LIVING-RELATED DONOR LOBECTOMY
Ann Thorac Surg 1994;5714234
DISCUSSION DR JOEL D. COOPER (St. Louis, MO): Doctor Cohen, that was really an excellent presentation, and you and your colleagues are to be greatly congratulated for an outstanding accomplishment, both in terms of the technical accomplishment and in terms of the early results you have achieved in your recipients. I want to start with some comments regarding technique. You have confirmed Santayama's dictum, in this case referring to the long-recognized problems peculiar to the right middle and lower bilobectomy, namely, an unusual tendency for the development of complications or of bronchial stump leaks or even empyemas. Such a stump leak virtually never occurs with any other type of lobectomy. Your complications in these patients, reflected by a very long postoperative stay, are not unique, and your switch to a right lower lobectomy, I believe, is quite appropriate. I question the need to heparinize the donor with its potential risk. In the early days of lung transplantation we heparinized neither lung donors nor recipients and did not even flush the lungs. After all, even removal of a lobe or a lung for cancer without any particular precautions does not lead to blood clot in the vessels. So I wonder, why don't you just isolate the hilar structures and then quickly excise the lobe and flush it on the back table? I note that if you do the calculation for all 14 donors, the average hospital stay was unusually long, namely, 12 days, 9 days if you exclude the bilobectomies, which is still quite long. Why was the postoperative stay so long? I think you probably addressed that in part. Were there any complications relating to the anticoagulation used? But clearly the most controversial aspect of the report is the use of living donors and the factors used to judge the risk-benefit ratio associated with this technique and its potential for a 300% mortality. The use of living donor lungs clearly has many appealing features: a normal lung, available when needed, elective timing, short ischemic time, daytime operation. Nonetheless, the success rate with lung transplantation at many centers, including our own, well exceeds 90%, even without these beneficial factors associated with a living donor, and this degree of success now does pertain to cystic fibrosis patients who have been on a ventilator anywhere from a week to a year and a half. And of course, as you know, there is no long-term immunologic benefit for using a parental lung; quite the contrary may be in fact the case. Therefore I would like to know how many of these recipients were on a ventilator before you decided to abandon the search for cadaver lungs or lobes and proceed with the use of living donor? If any of your recipients had not deteriorated to the point of ventilator dependence, why did you proceed? How many times have you used the lower lobes from adult cadaver donors for such bilateral transplants? What estimate of 5-year recipient survival did you give the donors in discussing this? Which raises my final question, namely, is the concept of informed consent limited or even abrogated when dealing with a parent-child relationship? After all, what parent would not risk life or limb to preserve a child's life? If a lobe today, maybe a whole lung tomorrow? What is the responsibility of the physician to restrict
the application of this type of procedure in spite of family or parental wish? I would be the first to admit that I have the questions but none of the answers. I do believe, in truth, that you and your team should be justifiably proud of this accomplishment, which if nothing else highlights the urgent need for more donors. DR COHEN: Doctor Cooper, thank you very much for your comments. I am particularly honored to have you as the discussant of this paper. The use of heparin in our living-related donors comes from its routine use in cadaveric donors. Although the amount of heparin that we give is much less than that used in cadavers, we want to make absolutely certain that thrombosis is prevented in the donor lung. This is particularly important because the donor lobe is removed from the donor and taken to a separate table before perfusion with the cold preservation solution is initiated. We also are careful to prevent thrombosis in the donor when we have to clamp proximal to remaining pulmonary branches, such as the right middle lobe artery, to repair these vessels without compromising them. There have been no complications as a result of heparin use; there have been no incidents of bleeding, and no transfusions in the donor group. Regarding our prolonged hospital stays, because of many of the ethical issues that you mentioned, we have been extremely conservative in our care of this first group of living donors. These patients have had to fulfill strict criteria regarding air leaks, drainage from chest tubes, and overall recovery until we have been convinced that they were really well enough to go home and were not going to return with further problems. As we become more comfortable with this procedure, our postoperative stays are improving, to the point where of the last 4 right lower lobe donors, 3 have left the hospital within 5 days. I would anticipate that our hospital stays will continue to improve as we get more and more experience with the procedure. In terms of the indications for living-related transplantation, those questions are very difficult to answer from an ethical standpoint. There were 2 patients who we believed were deteriorating rapidly and would not survive another 24 to 48 hours. One of them had a CO, tension of 140 mm Hg on the ventilator and was severely hypoxic such that we felt fortunate in just getting him to the operating room. We have, however, subsequently performed transplantation in some patients on a more semielective basis and, as you would expect, have found that the better their preoperative condition, the better their postoperative course. I would agree with you that at this point we cannot say that the long-term results of living-related lobar transplantation are comparable with those of cadaveric transplantation. We emphasize this to the families, especially the donors. However, it is a procedure that we have performed safely with good results up to this point. I agree with you that parents tend to be extremely enthusiastic about doing anything to help their children, and are more than agreeable to donate. I can only assure you that we are very cautious in our evaluations of these patients and are proceeding very carefully.