2000 Guideline
Based on strong evidence that patients with small call lung cancer (SCLC) who achieve a complete remission after induction therapy (chemotherapy or chemoradiotherapy) have a substantial decrease in the frequency of brain metastases and improved disease-free survival, and on an individual patient data meta-analysis demonstrating an improvement in overall survival, the Lung Disease Site Group (DSG) felt confident to recommend the use of prophylactic cranial irradiation (PCI) in this situation.
Meta-analyses provide an estimate of the overall magnitude of a treatment effect for a body of available evidence. However, meta-analyses should be carefully assessed before relying on them as the basis for a treatment recommendation in the absence of a large, definitive trial. Because of concern about the robustness of the meta-analysis, the modest value of p=0.01 and the absence of randomized trial data supporting PCI-improved overall survival, the authors of this practice guideline report and the members of the Lung DSG advise caution in the interpretation of these data. However, there is strong evidence from four published randomized controlled trials that PCI decreases the frequency of brain metastases and increases the disease-free survival rate in patients with SCLC who achieve complete responses after induction chemotherapy or chemoradiation therapy.
Members of the Lung DSG discussed whether those patients with extensive disease who achieve complete remission should receive PCI, particularly as their overall survival is shorter than the survival of patients with limited disease and there are virtually no long-term survivors. The Lung DSG concluded that, as the randomized trials included patients with extensive disease who had achieved complete response and there was an overall benefit in terms of disease-free survival, patients with extensive disease should not be denied the potential benefit of reduced risk of central nervous system metastases. This recommendation also takes into account the fact that staging of lung cancer may not be accurate and that the presence of extensive disease may not have been proven by biopsy. Although Lung DSG members acknowledged that central nervous system metastases can be treated when they occur, the psychological and physical consequences of brain metastases are grave, the neurological sequelae of the metastases often resolve incompletely after treatment, and survival is generally short once central nervous system metastases occur. For these reasons, the Lung DSG concluded that it was reasonable to offer PCI to those patients with extensive disease who achieve a complete response, in an effort to extend disease-free survival and maintain a good quality of life for the patient for as long as possible.
The widespread adoption of PCI following the achievement of a complete response to treatment in SCLC has been inhibited by concerns about acute and late neurological sequelae. These concerns arose from early small reports of acute neurological deterioration of cognitive and other neurological functions when radiotherapy was given in large fractions, high total dose and in combination with chemotherapy drugs, particularly nitrosoureas. Concerns about these potential neurological consequences and American concerns about medical-legal actions strongly influenced care providers against using PCI. Recent trials of PCI have carefully assessed the cognitive functioning of patients before, during, and after treatment. It is noteworthy that cognitive functioning, at least as measured by psychometric instruments, is commonly impaired prior to the administration of PCI and has shown no deterioration during PCI relative to those who do not receive PCI. Therefore, most investigators have concluded that serious acute neurotoxicity is not a major concern when the doses of radiotherapy recommended in this report are used. Although not commented on in the studies reviewed, Lung DSG members felt that short-term somnolence following PCI was a common side effect.
Most studies have not followed patients for more than several years. However, Lung DSG members have observed individual patients who have survived five or more years who have developed dementia. Whether this is directly related to PCI or to other factors is unknown, but it remains a concern and follow-up studies of long-term survivors are needed to inform this issue.
There is insufficient evidence available to comment on the optimal timing of PCI in relation to the administration of chemotherapy. The Lung DSG felt that PCI should not be concurrent with chemotherapy because of the potential interaction of the drugs and radiation on the brain vasculature or neural tissue, which might increase the risk of late neurotoxicity. Members of the Lung DSG felt that it should be given as soon as possible after completion of chemotherapy in complete responders.
In an attempt to obtain additional information about possible dose-response relationships, the Lung DSG considered a review of PCI versus no PCI for patients with SCLC which did not meet the inclusion criteria for this systematic review. Suwinski et al published a review based on a total of 40 trials of PCI versus no PCI, including 11 randomized and 12 nonrandomized trials, most of which involved patients with SCLC who had not achieved complete response to induction therapy. Two of the six randomized controlled trials cited above were included in the Suwinski analysis. The authors reported that a dose range of 30 to 35 Gy in 2 to 3 Gy fractions reduced the incidence of brain metastases within the remaining lifetime of the patients by 80%. They also suggested that PCI be administered early (within 60 days of starting induction therapy) which would mean that many patients would receive cranial irradiation concurrently with chemotherapy and would not have achieved complete remission. The toxicity resulting from concurrent administration of PCI and chemotherapy was not discussed.
Members of the Lung DSG expressed concern that too low a dose of radiation is probably ineffective. The radiation dose recommendation is 30 to 36 Gy in 2 to 3 Gy fractions or the biological equivalent.
2003 Update
As a result of feedback received during the peer review process prior to the publication of this guideline, the Lung DSG amended the description of Suwinski et al from "meta-analysis" to "review" and revised the last paragraph of this section as follows:
Members of the Lung DSG expressed concern that too low a dose of radiation is probably ineffective. The Lung DSG concluded that there is some indication that 30 to 36 Gy in 2 to 3 Gy per fraction or a biologically equivalent dose may produce a better outcome than a lower dose or less aggressive fractionation regimen.