Surveillance of Small-Cell Lung Cancer Patients Is Focus of Guidelines Debate

FORT LAUDERDALE, Fla--One of the most debated topics of the preliminarypractice guidelines prepared by the National Comprehensive CancerNetwork (NCCN) has been the level of surveillance after treatment,and this was also the case with the preliminary guideline forsmall-cell lung cancer (SCLC), presented by George D. Demetri,MD, at the NCCN's first annual conference.

The NCCN is a coalition of 15 major US cancer centers that havecome together to develop practice guidelines and perform outcomesresearch (see article).

In patients who have a complete response to initial therapy, theSCLC panel (see Table 1 for a list of the members) felt that frequentsurveillance is justified, since 5-year survival, although quitesmall, is not zero, said Dr. Demetri, of the Dana-Farber CancerInstitute.

A physician in the audience from the Fred Hutchinson Cancer Centersaid that surveillance in complete responders might be "overkill,since once the patient recurs, there isn't much that can be done."He suggested that physicians should be guided by the patient'ssymptoms.

Dr. Demetri acknowledged that this recommendation had been thesubject of much discussion among panel members, especially inlight of the breast cancer panel's attempt to pare down surveillance.

"But," he said, "we were struck with the fact thatif these patients were going to be palliated, the next level oftreatment would probably have to start before they had terriblesymptoms or liver dysfunction, which might preclude effectivepalliation, salvage, or entry into a clinical trial."

A researcher from the Mayo Clinic stated that an abstract to bepresented at the upcoming American Society of Clinical Oncology(ASCO) meeting will shed some light on whether such follow-upis useful in complete responders.

The study looked at the validity of physical examination, symptoms,CBC, chest x-ray, and chemistry panel in detecting relapse andalso at patient outcomes. Although the Mayo Clinic physician couldnot give detailed study results, she said that, basically, thestudy found that "none of these follow-ups were really helpful."

Dr. Demetri commented that once the Mayo data are published, theresults can be incorporated into the NCCN guidelines. However,based on his personal experience, he believes that surveillancein this situation can have an effect on a patient's quality oflife.

"It's very clear to me that if you pick up relapse earlier,before patients have end stage liver disease, for example, ordiffuse bone pain, or have lost 30 pounds, then they can havea much better quality of life, even though there is no survivaladvantage."

Few Sparks Fly in Presentation

Other than the surveillance issue, the presentation provoked littleheated debate, and, according to Dr. Demetri, there were few strongdisagreements among the panel in writing the guideline.

At the beginning of his talk, Dr. Demetri stressed that in SCLC,a disease with "wonderful response rates and dismal outcomes,"clinical trials are "justified out of the starting box."

Obviously, he said, no one is happy with the survival plateausat 5 years of 5% to 10%. "This is an ideal example of a diseasein which the NCCN can take the stance that well-designed, scientificallyrigorous clinical trials represent best standard care," hesaid, "so these guidelines were really designed for patientsfor whom clinical trials are not appropriate or are not available."

Since SCLC is a rapidly moving disease, the panel advocates arapid initial evaluation, to be done within 1 to 2 weeks. To accomplishthis, Dr. Demetri explained, the panel endorsed an evaluationpackage of tests rather than an algorithm where you "do apiece and then, if a certain finding is obtained, do another piece,and so on, because that typically could slow down the initialevaluation."

The initial evaluation package includes BUN, creatinine, and ECGto look for comorbidity "because of the predilection of thisdisease to occur in a slightly more advanced age group,"he added.

Although the preliminary guideline placed bone scan in the stagingphase, in the next version of the guideline it will be moved intothe initial evaluation. "We felt that trying to have bonescans as a se-quential study probably was not worth the possibledelay in ultimate treatment," he said.

If LDH is elevated in a patient with otherwise limited disease,the panel put bone marrow aspirate and biopsy as a Level 1 recommendation,while making this a Level 3 recommendation for the LDH negativepatient with otherwise limited disease. [See Table 2 for an explanationof the recommendation levels.]

"If the LDH is normal, a marrow aspirate is unlikely to bepositive, but it can be; you still might pick up 5% to 10% ofpatients and shift them from limited disease to extensive disease,"he said.

Treatment Recommendations

For the small subset of patients who have very limited resectabledisease (solitary pulmonary nodule), the panel recommends resectionand combination chemotherapy, but whether to also give radiationtherapy in this situation remains "a Level 3 or even Level4 disagreement," Dr. Demetri said.

For patients with limited disease in excess of solitary pulmonarynodule and good performance status, there was no disagreementon the appropriate treatment--combination chemotherapy with concurrentradiotherapy. The panel recommends cisplatin-etoposide for fourcycles (Level 2), but Dr. Demetri pointed out that other NCCNclinicians requested more flexibility in choosing a chemotherapyregimen.

"We will probably liberalize this to recognize the differentpractice patterns and then let individual institutions make thosedecisions on their own," Dr. Demetri said. "We may alsoliberalize the recommendation for four cycles to say, four tosix cycles."

Lower Dose Limit Recommended

As to radiation dose, the panel recommended only a lower limit(45 Gy/day). "At this time, we don't feel comfortable specifyingan upper limit. We felt that should be left to the different institutions,"he said.

For limited disease patients with poor performance status, sequentialrather than concurrent chemotherapy and radiotherapy is recommended.But, Dr. Demetri pointed out, "even in that subset we mightwant to give physicians the option to give concurrent radiotherapy.Poor performance status and the degree of comorbidity are oftenin the eye of the beholder."

For patients with extensive disease, the panel recommends etoposide/cisplatinalternating with CAV, but Dr. Demetri noted that some data favorifosfamide-based regimens. "We would like to see more studieson that. This gets back to my bias that clinical trials are goingto be the standard of care for this population."

For severely debilitated patients who have extensive stage disease,the guideline panel recommends a low intensity regimen such asoral etoposide, but if the debility is due to the cancer ratherthan other health problems, "one might choose to be moreaggressive," Dr. Demetri said. He noted that the next versionof the guideline will build more flexibility into that treatmentarm.

"This is why guidelines are not re-cipes. You can't algorithmout all the different vicissitudes of clinical practice. We aretrying to give some sort of a broad overview," he said.

No Surveillance for PRs, PDs

After the assessment of response to initial therapy, patientswith progressive disease (PD) go right to palliation, clinicaltrial, or salvage chemotherapy, Dr. Demetri said.

He noted that there was some Level 3 disagreement about whetherthese patients should receive any additional treatment at all."But some institutions wanted to offer one last chance tothese patients who are basically refractory to treatment at theoutset."

The partial responders (PRs) represent a more complicated subset,he said. The preliminary guideline recommends no surveillancefor these patients, who, like those with progressive disease,would proceed directly to palliation, salvage, or experimentaltreatment.

The next version of the guideline, he said, will suggest thatmonthly surveillance may be appropriate for these patients. "Thedisease can really move quickly once the relapse starts, and ifwe do anything shorter than monthly surveillance, by the timeit's caught, the patient may be too ill to benefit from palliationor to enter a clinical trial."

In patients judged to have a complete response to therapy, thepanel recommends that the physician and patient discuss the potentialrisks and benefits of prophylactic cranial irradiation.

"We could not say definitely, yes, all these patients shouldget it, or no, they shouldn't," he said. "But in anycase, it should be limited to patients with a complete responseor at last an extremely good partial response that's masqueradingas a near complete response."