Survival Impact of Prophylactic Cranial Irradiation in Limited-stage Small-cell Lung Cancer

Reviewer: Charles B. Simone, II, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: November 9, 2009

Presenter: M.E. Giuliani
Presenter's Affiliation: Princess Margaret Hospital, Toronto, ON, Canada
Type of Session: Scientific


  • Among patients with small-cell lung cancer (SCLC), approximately 10-15% are found to have brain metastasis at diagnosis. In total, approximately 30-60% of patients develop brain metastasis during the course of their disease. Among patients achieving a complete response to initial therapy, 50-80% will subsequently develop brain metastasis, with 20-30% developing intracranial metastasis as their sole site of metastasis, potentially suggesting the brain to be a tumor sanctuary site.
  • Following the recent publications of a survival benefit in randomized clinical trials and a meta-analysis, prophylactic cranial irradiation (PCI) is considered as the standard of care for patients with limited-stage SCLC with a complete or strong partial response to thoracic therapy and those with extensive-stage SCLC with an initial response to chemotherapy.
  • This study assessed the impact of PCI on overall survival and brain failure-free survival in patients with limited-stage SCLC and attempted to determine factors that impact the utilization of PCI.

Materials and Methods

  • Between 1997 and 2007, 796 patients were treated at Princess Margaret Hospital for SCLC.  Of these, 227 patients (28.5%) had limited-stage SCLC and were treated with radical therapy, with four to six cycles of chemotherapy with sequential or concurrent thoracic radiation therapy.   These patients formed the basis of this retrospective review.
  • Patients found to have progressive disease at restaging following chemoradiation did not receive PCI and were excluded from the analysis. Patients who did receive PCI were treated to 25 Gy in ten fractions in most cases. 
  • Study endpoints were overall survival and brain failure-free survival from the end of PCI therapy.


  • Of the 227 patients treated radically for limited-stage SCLC, 56% received PCI. The most common reasons for these patients not receiving PCI were patient refusal (30%, typically due to concerns about PCI toxicity), disease progression (22%), and patients lost to follow-up (15%). Younger patients and patients with a better performance status were statistically more likely to receive PCI, while there was no difference in rates of receiving PCI with gender. An insignificant increase in PCI was noted while time, with 51% of patients receiving PCI prior to 2001 and 61% thereafter. 
  • Of analyzable patients, 54% were male and their median age was 65 years. 
  • The median follow-up time was 16.7 months (range of 3 to 130 months).
  • Brain failure-free survival at 12, 24, 36, and 60 months was 77%, 41%, 41%, and 41%, respectively, for patients who did not receive PCI and 95%, 77%, 75%, and 69%, respectively, for patients who did receive PCI (p < 0.001).
  • Overall survival at 12, 24, 36, and 60 months was 74%, 35%, 27%, and 13%, respectively, for patients who did not receive PCI and 94%, 56%, 46%, and 33%, respectively, for patients who did (p < 0.002).  The overall survival for all limited-stage SCLC patients at 12, 24, 36, and 60 months was 86%, 48%, 39%, and 26%, respectively.

Author's Conclusions

  • Consistent with previous published findings, PCI was found in this study to significantly improve overall survival and brain failure-free survival in patients with limited-stage SCLC.
  • PCI should be considered for all patients with limited-stage SCLC. However, not all eligible patients are receiving PCI.
  • Talking with patients regarding their concerns of toxicity associated with PCI may allow for more patients to receive this intervention.

Clinical/Scientific Implications

  • This study confirmed recently published data demonstrated a survival benefit from PCI in patients in SCLC. While earlier smaller randomized trials mostly involving patients with limited stage SCLC have shown significant decreases in the rate of brain metastasis as the site of first relapse and overall rate of brain metastasis [Arriagada R, et al. J Natl Cancer Inst. 1995;87(3):183-90. and Arriagada R, et al. Ann Oncol. 2002;13(5):748-54.], no survival benefit with PCI was definitively demonstrated. In 1999, a meta-analysis of seven trials with 987 patients with SCLC in complete response were randomized from 1965-1995 to PCI versus no PCI. In that trial, PCI decreased the incidence of brain metastasis by 25.3% (33.3% vs. 58.6%, RR 0.46, p<0.001) and increased the 3-year overall survival by 5.4% (20.7% vs. 15.3%, RR 0.84, p=0.01) and disease-free survival by 8.8% (22.3% vs. 13.5%, RR 0.75, p<0.001) [Aupérin A, et al. N Engl J Med. 1999;341(7):476-84.]. More recently, a similar survival benefit with PCI has also been demonstrated in patients with extensive stage SCLC [Slotman B, et al. N Engl J Med. 2007;357(7):664-72.].
  • No information is available regarding the timing of PCI in this single-institution retrospective patient population. Based on previously published data, the benefit of PCI is greatest if initiated within six months of initiating chemo and within four to 12 wks after completing chemotherapy.
  • As this study cohort includes patients treated before the recently published randomized Intergroup trial of 720 patients with limited stage SCLC in complete remission after chemotherapy and thoracic radiation RT that demonstrated the superiority of PCI doses of 25 Gy in ten fractions over 36 Gy in 18 to 24 fractions (Le Péchoux C, et al. Lancet Oncol. 2009;10(5):467-74.), not all patients in this study were treated with this fractionation regimen. Current recommendations advocate a PCI dose of 25 Gy in 10 fractions.
  • PCI is often withheld from patients with multiple comorbidities, poorer performance status, or baseline impaired mental function. However, many patients eligible for PCI did not receive this therapy in the current study. Concerns of neurological toxicity may have contributed to patients not pursuing PCI and practitioners not recommended this intervention. Neurological toxicity following PCI has been demonstrated to be higher with fractions sizes greater than 3 Gy or with concurrent chemotherapy, neither of which were performed in this patient cohort. However, older studies have demonstrated an increased rate of neurocognitive decline with PCI in elderly patients when compared to younger patients. This may account for the findings in this study that fewer elderly patients received PCI.
  • Newer studies do not clearly demonstrate a significant neurocognitive decline attributable to PCI, and several studies instead suggest that pre-existing neurotoxicity may contribute to neurological morbidity observed following PCI. Etiologies of pre-PCI morbidity include paraneoplastic limbic encephalopathy, chemotherapy, and occult intracranial micrometastasis.
  • PCI has convincingly been demonstrated to increase overall survival and brain metastasis-free survival in patients with SCLC. The study confirms the importance of need for practitioners to discuss PCI with all of their patients with SCLC, and to have balances discussions regarding the risks and benefits of this intervention.