The search for effective postoperative adjuvant therapy for patients with resected non-small-cell lung cancer (NSCLC) has been spurred by a high rate of failure after definitive surgery. Except for patients with resected T1, N0, M0 lesions, failure rates exceed 30%. Widespread application of adjuvant therapy has been reined in by a disappointing lack of effectiveness in this setting.
The search for effective postoperative adjuvant therapyfor patients with resected non-small-cell lung cancer (NSCLC) has been spurredby a high rate of failure after definitive surgery. Except for patients withresected T1, N0, M0 lesions, failure rates exceed 30%. Widespread application ofadjuvant therapy has been reined in by a disappointing lack of effectiveness inthis setting.
The article by Dr. Movsas provides an evenhanded review of some of the keycontroversies currently bedeviling the clinician faced with making arecommendation for the patient with resected node-positive NSCLC. The article isgratifying in its desire to stay close to the evidence rather than join thepolemics that often surround such discussions, and it largely succeeds in thisgoal. What is disappointing is that the present data do not allow the author, ora fair-minded reader, to make any firm conclusions and treatmentrecommendations.
How is it that we know so little about optimal adjuvant therapy for a cancerthat each year kills more people in the United States than cancers of thebreast, prostate, and colon combined? What has brought us to this sorry statewhere medical and radiation oncologists spend time disputing meta-analyses ofdecade-old data, and how can we move off this sticking point in the next decade?
Large Adjuvant Trials Lacking
There has been a tendency in designing adjuvant trials to think that ourtherapeutic interventions are likely to be of great benefit, improving overallor relapse-free survival by 10% or more. It is highly unlikely that this will bethe case, and a lesser degree of benefit (eg, an absolute gain of 5% insurvival) would be considered clinically worthwhile by many patients andphysicians.[1,2]
Even the demonstration of such small differences, however, requires largetrials, and until relatively recently, there has been a reluctance to implementsuch large studies, at least in the United States. Several European trials ofadjuvant therapy have recently been completed, with relatively broad eligibilityrequirements and flexibility in the choice of chemotherapeutic regimen. Theresults of these trials are awaited with great interest. At best, however, theywill give provisional answers about the value of what is rapidly becomingyesterday’s chemotherapy. We have probably missed the window in which amodestly but significantly effective adjuvant chemotherapeutic regimen couldbecome a broad-spectrum standard analogous to what fluorouracil (5-FU)/leucovorinin colorectal cancer or CAF (cyclophosphamide, doxorubicin [Adriamycin], 5-FU)had been in breast cancer 5 years ago. We are approaching a time whentherapeutic recommendations will be much more individualized.
Another course of action is to develop tools that better define patients athighest risk for relapse and to develop adjuvant therapies targeted morespecifically against the molecular features of their tumors. We are rapidlyapproaching the time when such strategies will be feasible in the clinic.
A combination of the sentinel node method for identifying the pattern ofdrainage of individual tumors and immunohistochemical staining and thin-sectionanalysis can allow us to identify otherwise occult lymph node involvement in asubstantial proportion of patients with breast cancer and malignant melanoma.Early results suggest that these techniques will be applicable in lung cancer aswell.[3-5] Failure to detect any such occult spread should identify a group ofpatients at truly low risk for relapse who may be spared adjuvant therapies.
New Systemic Approaches
While modest changes in the chemotherapeutic regimens now employed will beplayed out over the next few years for metastatic disease and in the adjuvantsetting, it is unrealistic to expect that there will be large differences inoutcome. The paradigm of conventional nonspecific cytotoxic agents has not beenhighly effective in metastatic NSCLC. We are, however, at the threshold ofusing new classes of agents that inhibit, in relatively specific fashion, keyenzymes involved in the neoplastic process. These include ras prenylation, avariety of receptor tyrosine kinases (with both monoclonal antibodies developedagainst the receptor moiety for HER2 [aka c-erbB-2] and epidermal growth factor[EGFR] and low-molecular-weight ATP binding site inhibitors of the kinasemoiety), cyclooxygenase-2, and matrix metalloproteinases. To date these agentshave shown clinical promise in metastatic disease and relatively mild toxicity.
How Important Is Local Control?
The role of adjuvant radiation therapy is likely to remain uncertain. Much ofthe past controversy has been generated because of poor choice of patients (eg,resected N0 patients at low risk of relapse) and technically poor radiationtherapy and choice of fractionation schedules with a high risk of late adverseeffects. There are solid reasons to expect reduced morbidity with moreappropriately delivered radiation therapy based on CT planning and increasinguse of conformal and intensity-modulated radiation therapy to limit thehigh-dose volume to the lymph node regions at risk.
The question in the coming decade, as systemic adjuvant therapies are morewidely used, will be whether there exists a group of patients with asufficiently high local relapse rate, despite adjuvant systemic therapy, forwhom adjuvant local therapy is still warranted. In this setting, we can properlyassess whether such therapy will improve local control (almost certainly),quality of life (probably), and overall survival (my guess is yes, but I wouldcertainly look to a randomized trial on this question).
If this sounds a bit familiar, it is. Those of us with a touch of gray recallthe same arguments being played out in the breast cancer arena 2 decades ago,with eerie similarities concerning the supposed harm of adjuvant radiation. Somesay that "the mediastinum is the axilla of the lung"a widespreadobservation with more than a grain of truth to it.
Prophylactic Cranial Irradiation
An additional area for consideration of adjuvant therapy of NSCLC is theproblem of central nervous system (CNS) relapse. As we have become modestly moresuccessful in extending survival for patients with stage IIIA (N2) disease,several groups have reported that the actuarial risk of CNS relapse in NSCLCapproaches 60%, not much different from that seen in small-cell lung cancer(SCLC).[6-8]
Although several trials of prophylactic cranial irradiation in unselectedpatients with NSCLC did not show any clinical benefit, more recent reportssuggest that, in groups of patients at high risk for CNS failure (stage IIIA,N2), modest doses of prophylactic cranial irradiation have similar effectivenessin NSCLC and SCLC. Hopefully, the true value of such treatment can beevaluated in a small number of adequately powered phase III trials and not waitfor 2 decades and a meta-analysis of many small underpowered trials.
Our track record in thoracic oncology has not been encouraging when it comesto accruing patients to randomized trials of adjuvant therapy. It has beenparticularly difficult in past years, with the sometimes emotionally overwroughtclaims of the worthlessness or proven efficacy of particular modalities, whensuch claims rest on significantly flawed meta-analyses.
We face a time of accelerating change in our ability to target tumors, atboth macroscopic and molecular levels. We also have the potential to develop andevaluate adjuvant approaches far more sophisticated than the ones we have beenso excessively meta-analyzing. It is time to stop squabbling over the past andlook for new therapies that work.
1. Yellan SG, Cella DF: Someone to live for: Social well being, parenthoodstates, and decision making in oncology. J Clin Oncol 13:1255-1264, 1995.
2. Leighl N, Gattellari M, Butow P, et al: Discussing adjuvant cancertherapy. J Clin Oncol 19:1768-1778, 2001.
3. Liptay MJ, Masters GA, Winchester DJ, et al: Intraoperative radioisotopesentinel lymph node mapping in non-small-cell lung cancer. Ann Thorac Surg70(2):384-390 (incl discussion), 2000.
4. Little AG, DeHoyos A, Kirgan DM, et al: Intraoperative lymphatic mappingfor non-small-cell lung cancer: The sentinel node technique. J ThoracCardiovasc Surg 117(2):220-234, 1999.
5. Naruke T, Tsuchiya R, Kondo H, et al: Lymph node sampling in lung cancer:How should it be done? Eur J Cardiothorac Surg 16(suppl 1):S17-S24, 1999.
6. Breathnach OS, Boris Freidlin B, Conley B, et al: Twenty-two years ofphase III trials for patients with advanced non-small-cell lung cancer:Sobering results. J Clin Oncol 19:1734-1742, 2001.
7. Law A, Daly B, Madsen M, et al: High incidence of isolated brainmetastases (CNS Mets) following complete response (CR) in advanced non-small-celllung cancer (NSCLC): A new challenge (abstract). Proc Am Soc Clin Oncol16:A1604, 1997.
8. Robnett TJ, Machtay M, Stevenson JP, et al: Factors affecting the risk ofbrain metastases after definitive chemoradiation for locally advanced non-small-celllung carcinoma. J Clin Oncol 19:1344-1349, 2001.
9. Stuschke M, Eberhardt W, Pöttgen C, et al: Prophylactic cranialirradiation in locally advanced non-small-cell lung cancer after multimodalitytreatment: Long-term follow-up and investigations of late neuropsychologiceffects. J Clin Oncol 17:2700-2709, 2000.