| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
Yale-New Haven Hospital New Haven, CT, ; Dr. Ponn is Assistant Clinical Professor and Associate Section Chief, Cardiothoracic Surgery, Yale-New Haven Hospital and the Hospital of St. Raphael.
Correspondence to: Ronald M. Ponn, MD, FCCP, Cardiothoracic Surgery, Yale-New Haven Hospital and the Hospital of St. Raphael, 330 Orchard St, New Haven, CT 06511
Most patients with non-small cell lung cancer (NSCLC) present at an advanced stage and are incurable by current treatment modalities. A smaller group, 20 to 25% of cases, in whom cancer is detected while it is still localized, have a reasonable chance of long-term survival. A majority of patients with stage I and a significant minority of those with stage II NSCLC achieve cancer-free survival for > 5 years after complete resection. Recurrence of the primary tumor is rarely amenable to curative therapy. In contrast, patients with localized second primary lung cancers who are physiologically able to undergo complete resection may benefit from reoperation.
In this issue of CHEST (see page 1621), Asaph and associates detail their experiences with the surgical treatment of patients who have developed a metachronous NSCLC. The authors identified 37 such cases from a database of about 800 patients who had undergone resection over the past 2 decades. Four of the patients had initial small cell cancers. The majority of initial cancers were stage IA lesions (78%), but 57% of second cancers were more advanced. Anatomic resections, including 10 completion pneumonectomies, were carried out in the majority of second operations (86%) and were associated with a low operative mortality (5.4%). The complication rates and postoperative lengths of hospitalization were likewise low. Including death from all causes, the 5-year and 10-year survival rates following resection of the second tumors were 33% and 10%, respectively. The majority of both first (45%) and second (60%) cancers were adenocarcinomas. Also, reflective of the changing epidemiology of lung cancer is that the report is unique in that the majority (60%) of the patients were women. Although the report is otherwise similar to other studies and is limited in that the total number of second lesions is unknown, since only operated cases were identifiable, it focuses our attention on this important clinical area. Among the many issues relevant to second lung cancers are their differentiation from metastasis, their incidence and changing histology, optimal therapy and prognosis, and the approach to surveillance following presumed curative treatment of the lung cancer patient.
The current authors, most previous reports, and clinicians in daily practice apply the criteria delineated by Martini and Melamed1 for defining metachronous lung cancers. Although some have suggested that a disease-free interval of 3 years or 4 years be required,2 3 most accept the 2-year standard. It must be stressed, however, that confident differentiation of a new primary from a metastasis is often impossible. Even a difference in histology is not an absolute indicator, since many tumors are comprised of variable cellular elements, whether defined by degree of differentiation or by actual cell type, and a metastasis may consist of only one of the clones present in the initial lesion. Compounding the dilemma is the common finding of identical histology in the first and subsequent cancers, 35% in the present series, and from 50 to 75% in other reports.2 3 4 The same problem relates to synchronous lesions. In some sense, the designations synchronous and metachronous are arbitrary, especially in cases with short intervals, since calculation of tumor doubling times indicates that most cancers are present for a number of years before becoming clinically detectable. Although not commonly used, additional evidence can be gleaned from genetic studies. In small series, applying DNA flow cytometry5 or analysis of p53 gene mutations,4 most lesions diagnosed by clinical criteria as separate primaries are found to be discordant from the original tumor. Although this suggests that the clinical criteria are largely accurate, it must be kept in mind that the polyclonality of many tumors indicates that some will inevitably be misclassified. Although given our current therapeutic options, the treatment plan is not altered in most cases. Results reporting and clarification of natural history would clearly benefit from more definitive data.
Operations for second lung cancers made up about 4.5% of pulmonary resections for cancer in the present series. In other series, this rate varies from 0.8% to as much as 10%.4 6 7 8 Since it is limited to surgical patients, the present series does not allow calculation of the more important rate of development of NSCLC after an index cancer. It has been clear, however, ever since long-term data on sufficient numbers of treated patients became available > 2 decades ago,1 9 that these patients are at high risk for multiple types of aerodigestive second tumors, especially second lung cancers. This risk has been confirmed in later cooperative and large institutional studies6 10 11 and by updates of the original reports.12 The approximate rate of a second primary lung cancer for a curatively resected NSCLC patient is about 1 to 2% per patient per year,13 has remained steady over time, and does not appear to decrease significantly with the passage of time. The histology of second lung cancers, however, has changed from overwhelmingly squamous cancers to adenocarcinomas. This change is reflected in the present series, as well as other recent reports.14 15 16 A particularly high rate of 5 to 6% per patient per year has been noted in two smaller groups of patients. As many as 45% of surgically treated cases of "occult" lung cancer develop second primaries during follow-up,17 but this entity is rarely encountered currently. More likely to become an increasing problem are patients successfully treated for limited stage small cell carcinoma, in whom the risk of a metachronous NSCLC may be as high as 69% 16 years after treatment.18
As for initial lung cancers, the optimal treatment for second tumors is complete resection. Five-year survival in about a third of cases, as noted in the present series, is in line with most reported experience, with some reporting > 40% in stage I cases.2 8 11 19 Operative mortality is low and remains low in the small group treated by resection for third primary lung cancers. It is obvious that compulsive evaluation is mandatory to ensure adequate pulmonary reserve and the absence of metastatic disease. It is also clear from the current report and others that, when necessary, completion pneumonectomy can be performed with acceptable results. Of necessity, second operations must sometimes be limited to segmental or wedge resection, but they should always be accompanied by interlobar and mediastinal nodal staging. The major prognostic factors for second cancers are the same as for index cancers, namely disease stage and completeness of resection.2 11 19 20 The late survival for second cancers, however, is lower than that for similar stage first tumors. The reason for this discordance is unknown and is likely multifactorial. In addition to the fact that patients with second lesions are older, one may speculate that the occurrence of a second primary represents an unfavorable biologic indicator or that resections are limited by technical and physiologic factors. It is also likely that the lower survival may be due to the currently unavoidable problem of admixing metastatic disease with true second primaries. In this regard, it is noteworthy that several authors have determined that a longer disease-free interval between the index and second cancer is associated with an improved prognosis.6 19 21 It is likely that more of these tumors are true second primaries. Survival is not influenced by cell type or whether the first and second cancers are of similar or different histology.11 16
The persistent risk of second cancers and the success of their surgical treatment, as well as the significant rate of recurrent disease, suggest that careful follow-up of the curatively treated lung cancer patient should be a worthwhile, lifelong endeavor. The conventional wisdom is that, for the first 2 or 3 years, we are looking for recurrence, while later new cancers are the issue. However, there exists no consensus as to how best to structure follow-up. It is disturbing that Asaph and colleagues are excluded from long-term follow-up of their patients because of managed care. This may explain the fact that all patients with second cancers in this series underwent surgery, in contrast to the 60 to 70% operability rate reported by others.1 6 8 11 One fears that, when medical and surgical thoracic oncologists are kept out of the surveillance and decision-making loop with respect to defining second primaries and how best to treat them, some patients may be denied a chance for cure. Although the efficacy of postoperative surveillance has been questioned,22 23 cogent counterarguments have been presented24 25 26 based on both clinical and humanistic benefits to the patient and the knowledge base and skills of the following physicians. Although a 1 to 2% risk of second cancer may be interpreted as small by a health-care administrator, the cumulative risk should be alarming to both patients and physicians—alarming in a positive way, causing the former group to abandon tobacco and the latter group to strive for better means of identification, definition, and treatment. If surveillance has been found wanting, it is likely the method that is at fault, rather than the theory. Although routine imaging other than chest radiography traditionally has been discouraged, the authors correctly point out that it may be time to reassess this belief. In light of the recently reported clinical and economic utility of low-dose CT scan for identifying early-stage cancers in smokers, it is logical to apply this approach in the posttreatment patient population, a group at 10 times the risk of smokers in general. Finally, the most desirable approach would be prevention. There is evidence that smoking cessation reduces the risk. Chemoprevention data are as yet inconclusive.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
