Greater Node Sampling Urged in Lung Surgery

FigureThoracic surgeons at major cancer centers are urging their colleagues at community hospitals to follow their lead in lymph-node sampling. "Many colleagues who do complete lobectomies only once in a while don't have the interest to do lymph-node sampling," said Steven Keller, MD, Clinical Professor of Cardiovascular and Thoracic Surgery at Albert Einstein College of Medicine and Co-Director of the Cardiothoracic Surgery Residency Program at Montefiore Medical Center. "Those of us who do only thoracic surgery will generally remove the nodes." Keller and other advocates for rigorous sampling warn that surgeons who remove too few nodes hamper staging. That, in turn, could delay chemotherapy that might benefit stage II or III patients. "The more lymph nodes you remove," Keller said, "the more accurately you stage the patient." Frank Detterbeck, MD, Professor of Surgery and Chief of Thoracic Surgery at Yale Cancer Center, concurs. Stage II and III, ideally, should get chemotherapy. You do have a higher chance of curing them." Eric Vallieres, MD, Surgical Director for Lung Cancer at the Swedish Cancer Institute in Seattle, says that inadequate node sampling forces oncologists into guesswork: "It's very hard for an oncologist to make a recommendation for or against chemotherapy when he doesn't have information." Another argument for extensive node sampling is that it's normally required for enrolling patients in clinical trials: "Many patients are not eligible for trials because there are not sufficient [lymph-node] levels sampled or dissected," Keller noted.FRANK DETTERBECK, MD, said that some surgeons wrongly assume that all nodal cancers will show up in a PET scan. "PET scanning is not a very good way to assess the mediastinal. The false-negative rate is pretty high."STEVEN KELLER, MD, warned that surgeons who remove too few nodes hamper staging, which, in turn, could delay chemotherapy that might benefit stage II or III patients. "The more lymph nodes you remove, the more accurately you stage the patient."Still another reason for lymphadenectomy, advocates say, is a potential therapeutic benefit for higher-stage patients. "I think the data at least for early-stage lung cancer is no," said Detterbeck. "Stage II and III, we don't know." Recommended Stations for Sampling Although some studies recommend a certain number of nodes to be sampled, the surgeons interviewed for this article said they base their recommendations on stations sampled. The recommended mediastinal stations for sampling in right-sided cancers are 2R, 4R, 7, 8, and 9. For left-sided cancers, the stations are 4L, 5, 6, 7, 8 and 9. Systematic sampling involves taking at least one node from each recommended station. A study published last year by the American College of Surgeons Oncology Group, based on data from 1999 to 2004 (J Thorac Cardiovasc Surg 2011;141:662–670), revealed occult N2 disease in the nodes of approximately four percent of 1,111 lobectomy patients. Ninety percent of them had at least 10 nodes taken from three stations. "The important thing is that one or more nodes are sampled systematically from standard lymph-node stations," said the lead author, Gail Darling, MD, Director of Clinical Research for Thoracic Surgery at Toronto General Hospital and Professor of Thoracic Surgery at the University of Toronto. "Generally, we would recommend at least three mediastinal lymph node stations including the subcarinal node, as well as the hilar lymph node and intralobar nodes—so this translates to a minimum of six nodes, but 10 or more would be better." The study showed no improvements in survival from going beyond systematic sampling to complete mediastinal dissection—removing all nodes from the recommended stations. "There's pretty good evidence that in terms of staging a patient accurately, whether you do systematic sampling or complete node dissection, you pretty much end up with the same stage," Keller said. Data from the NCI's Surveillance, Epidemiology and End Results (SEERS) database also suggests that systematic sampling is as beneficial as mediastinal dissection. Between 1990 and 2000, some 16,800 patients with non-small-cell lung cancer showed a modest improvement in survival when five to eight lymph nodes were examined (Chest 2005;128:1545–1550). Evaluating more than 16 nodes had no additional benefit. Yet, Keller and the other surgeons interviewed routinely perform complete mediastinal dissections in their lobectomy cancer patients, saying that they feel there are too few studies to disprove a benefit in later-stage patients and that morbidity from dissection is low. "I feel as if complete mediastinal dissection is better sampling," he said.GAIL DARLING, MD: "Generally, we would recommend at least three mediastinal lymph node stations including the subcarinal node, as well as the hilar lymph node and intralobar nodes—so this translates to a minimum of six nodes, but 10 or more would be better."Incentives Lacking With no enforced standards on node removal, many surgeons have little incentive to sample beyond nodes that look suspicious. "It adds to the time and complexity of the surgery," said Heather Wakelee, MD, Assistant Professor of Medicine, Oncology, at Stanford University and a thoracic oncologist at Stanford Clinical Cancer Center. "It's easier not to add that component. Surgeons are not thinking like cancer doctors." Guidelines from the National Cancer Institute and National Comprehensive Cancer Network lack teeth, those interviewed for this article said. Unlike guidelines on colorectal nodes, those for lung nodes typically are not used in evaluating hospital quality. Nor do insurers cover node removal during lobectomy. Surgeons estimate that systematic sampling adds 15 to 20 minutes to the surgery. There is also a belief among thoracic surgeons that most of the surgeons who eschew node sampling are general or heart surgeons practicing in community hospitals. "A lot of thoracic surgery is done by general and cardiac surgeons," Detterbeck said. "They're doing [lobectomies] sporadically. I don't think they do that good a job." A study on patterns of surgical care published in 2005 by the American College of Surgeons Commission on Cancer (Ann Thorac Surg 2005;80:2051–2056) showed that node biopsies indeed vary by setting. While nearly 52% of patients in teaching hospitals had lymph-node biopsies, the practice dropped to 45% in comprehensive community cancer centers and to 40% in community cancer centers. Some surgeons wrongly assume that all nodal cancers will show up in a PET scan, said Detterbeck. "PET scanning is not a very good way to assess the mediastinal. The false-negative rate is pretty high." Needed: Quality Metrics He and others have called for enforceable guidelines for node sampling, noting, though, that a major hurdle is deciding just what is optimal sampling. "We need quality metrics. It's difficult to assign a number of nodes. We need to come up with an appropriate quality measure," Detterbeck said. Until guidelines are enforced, he and other advocates recommend pressuring reluctant surgeons to sample systematically. "Pulmonologists and oncologists should expect at least one lymph node to be assessed from three mediastinal node stations," Darling said, "and if they are not getting that information, they should go back to their surgeon and ask, ‘What about the mediastinal nodes?’"HEATHER WAKELEE, MD, notes that with no enforced standards on node removal, many surgeons have little incentive to sample beyond nodes that look suspicious. "It adds to the time and complexity of the surgery. It's easier not to add that component. Surgeons are not thinking like cancer doctors," she said.Keller suggests a business approach: "Most surgeons are like small-business owners – They respond to their customers. You need to get on the phone and say, ‘Next time I need lymph nodes in the report or I'll send my patients elsewhere.’"

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.