Fax: (416) 946-6566
Utilization of prophylactic cranial irradiation in patients with limited stage small cell lung carcinoma
Version of Record online: 27 AUG 2010
Copyright © 2010 American Cancer Society
Volume 116, Issue 24, pages 5694–5699, 15 December 2010
How to Cite
Giuliani, M., Sun, A., Bezjak, A., Ma, C., Le, L. W., Brade, A., Cho, J., Leighl, N. B., Shepherd, F. A. and Hope, A. J. (2010), Utilization of prophylactic cranial irradiation in patients with limited stage small cell lung carcinoma. Cancer, 116: 5694–5699. doi: 10.1002/cncr.25341
- Issue online: 3 DEC 2010
- Version of Record online: 27 AUG 2010
- Manuscript Accepted: 1 MAR 2010
- Manuscript Revised: 25 FEB 2010
- Manuscript Received: 17 JAN 2010
- small cell;
- limited stage;
- prophylactic cranial irradiation;
- patient preference;
This study reports the adoption of prophylactic cranial irradiation (PCI) in patients with limited stage small cell lung carcinoma (LS-SCLC) at Princess Margaret Hospital (PMH) and the factors that impact PCI utilization.
A retrospective review was performed on all patients with LS-SCLC treated at PMH from 1997 to 2007. Clinical details including the rate of PCI utilization were determined and, for patients not receiving PCI, the documented reason was recorded. Brain failure free survival (FFS) and overall survival (OS) were estimated by the Kaplan-Meier method, comparing patients treated with or without PCI. Pearson chi-square test was used to determine factors associated with PCI use.
Two hundred seven patients were treated for LS-SCLC and 61.4% (n = 127) of these patients received PCI. The most common documented reason for not receiving PCI was patient refusal, typically because of concerns about PCI toxicity. Patients older than 65 were significantly less likely to receive PCI. Brain FFS and OS rates were significantly higher in patients who received PCI.
Not all eligible patients are receiving PCI, despite its significant effect on reducing brain metastases and improving OS. Emphasizing the benefits of PCI to patients, when discussing potential toxicities, may improve utilization. Cancer 2010. © 2010 American Cancer Society.
Small cell lung carcinoma (SCLC) accounts for 13% of all lung cancers, with 40% of cases presenting with limited stage disease.1 Presently, evidence supports management of limited stage small cell lung carcinoma (LS-SCLC) with platinum-based chemotherapy and concurrent thoracic radiotherapy,2 followed by prophylactic cranial irradiation (PCI).3, 4
In patients with SCLC, the cumulative probability of metastasis to the brain is 36% and 47% at 12 and 24 months, respectively, without PCI.5 Meta-analyses have shown that PCI decreases the incidence of brain metastasis and improves overall survival in patients with LS-SCLC.3, 4 PCI reduces the incidence of brain metastasis at 3 years from 58.6% to 33.3%, an absolute reduction of 25.3%, and improves overall survival by 5.4% at 3 years.4 PCI also has been shown to improve the median overall survival in patients with extensive-stage (ES) SCLC from 5.4 months to 6.7 months and reduce the incidence of symptomatic brain metastasis at 1 year from 40.4% to 14.6%.6
The purpose of this study was to assess the utilization of PCI in patients with LS-SCLC at Princess Margaret Hospital (PMH) in Toronto, assess the factors impacting PCI utilization, and assess the impact of PCI on brain failure free survival (FFS) and overall survival (OS).
MATERIALS AND METHODS
From 1997 to 2007, 796 patients with SCLC were treated at PMH and were identified via review of patient records. A retrospective chart review was conducted in accordance with hospital research ethics board approval. Data were collected on date of diagnosis, sex, age at diagnosis, histology, clinical stage, thoracic radiotherapy, prophylactic cranial irradiation, the presence and type of chemotherapy, smoking status, details of tumor recurrence, presence or absence of salvage treatment, adverse effects, and survival.
The patients who received chemoradiotherapy with curative intent for LS-SCLC were identified. Patients with progressive disease at restaging after chemotherapy and thoracic radiotherapy, who were not eligible to receive PCI per institutional practice at that time, were excluded (Fig. 1).
The majority of patients were staged with computed tomography (CT) of the thorax and abdomen, a bone scan, and a CT of the head or magnetic resonance imaging (MRI) of the brain before starting chemoradiation. Most patients were treated with cisplatin and etoposide chemotherapy, given sequentially or concurrently with radiotherapy. Thoracic radiotherapy was hypofractionated, 4005 cGy in 15 fractions, in the majority of patients as per Canadian practice.7 Thoracic radiotherapy was planned on 3-dimensional (3D) CT and mostly delivered in a 2-phase plan, initially anterior-posterior followed by an off-cord boost. Most patients were restaged with CT thorax/abdomen and CT of the head or MRI of the brain after completion of chemoradiation to ensure eligibility for PCI. In accordance with institutional policies, PCI was discussed with all patients except those who were lost to follow-up. Patients who received PCI were generally treated with 2500 cGy in 10 fractions. However, some patients were treated on clinical trial protocols and received different dose/fractionation schedules.
Patient demographic and disease characteristics were summarized using descriptive statistics. Pearson chi-square test was used to determine factors associated with PCI use. Overall survival (OS) was calculated from the end of chemoradiation to date of death or censored at the date of last follow-up (if alive). Brain failure-free survival (FFS) was calculated from the end of chemoradiation to date of brain failure or censored at date of death or censored at last follow-up date (if without brain metastases). Patients lost to follow-up were assumed not to have developed brain metastases. Brain FFS and OS curves were constructed by using Kaplan-Meier method. Log-rank test was performed to compare brain FFS and OS for patients treated with or without PCI. Multivariate analyses for brain FFS and OS were then performed using Cox proportional hazards model with stepwise variable selection procedure. All tests were 2-sided, and the statistical significance level was set at .05.
Chemoradiotherapy, with curative intent, was delivered to 228 patients with LS-SCLC (29% of all patients with SCLC). Twenty-one patients were excluded from the analysis as they had progressive disease at restaging after chemotherapy and thoracic radiotherapy and were not eligible to receive PCI per institutional practice at that time. Of these 21 patients, 3 (14.3%) had locoregional progression as the site of first progression and 18 (85.7%) had distant progression; 11 of those had developed brain metastases. The remaining 207 patients were analyzed. The median follow-up at time of analysis was 18.8 months, (range, 2.2 mo to 130.1 mo). The median follow-up of the patients alive at the time of analysis was 26.5 months. The median follow-up for patients who received PCI was 18.4 months versus 8.9 months for those who did not receive PCI. The median age was 65.7 years (range, 38.6 y to 88.1 y). Patient demographics are described in Table 1.
|Date of diagnosis|
A total of 127 (61.4%) of the 207 patients received PCI. There was a nonsignificant increase (from 57.9% to 66.0%; P = .21) in the use of PCI in patients diagnosed after the year 2001 versus those diagnosed on or before 2001.
Of the patients who did not receive PCI, 37.5% (30) refused the treatment, 18.8% (15) were lost to follow-up, 15.0% (12) had a partial response to chemoradiotherapy, 8.8% (7) had other reasons for not receiving PCI, and in 20.0% (16), the reason was not documented.
Of the 30 patients who declined PCI, 16 (53.3%) did so because of concerns about potential toxicity and 6 (20.0%) did so because of the toxicity they had already experienced with their chemotherapy and thoracic radiotherapy. In 8 (26.7%) of patients the reason was not specified. Fifteen patients did not receive PCI because they were lost to follow-up. Many of these patients were receiving chemotherapy at an outside hospital and did not return for consultation about PCI.
Patients older than 65 years of age were significantly less likely to receive PCI than patients aged 65 years or less; 45.7% versus 54.3% (P = .01). There was no statistically significant difference in PCI use based on sex (Table 2).
|PCI n=127||No PCI n=80||P|
|Men||69 (54.3%)||42 (52.5%)|
|Women||58 (45.7%)||38 (47.5%)|
|≤65||69 (54.3%)||29 (36.3%)|
|>65||58 (45.7%)||51 (63.7%)|
Brain FFS was significantly longer in patients who received PCI. The hazard ratio (HR) was 3.4 (95%, CI, 1.9-6.1; P < .001) for the patients who did not receive PCI versus those who did. At 3 years from completion of treatment, the proportion of patients alive without brain failure was 76.6% (95%, CI, 68-87) versus 46.7% (95%, CI, 8-34) for patients who did and did not receive PCI, respectively (Fig. 2). PCI treatment was associated with significantly longer brain FFS in the multivariate analysis after adjustments for age (P < .0001). The HR was 3.8 (95%, CI, 2.1-6.8) for the patients who did not receive PCI versus those that did.
Overall survival was significantly longer in patients who received PCI. The HR was 2.1 (95% CI, 1.5-3.0; P < .001) for the patients who did not receive PCI versus those that did. Estimated overall survival rates at 3 and 5 years from completion of treatment were 40.1% (95% CI, 32-51) versus 16.0% (95% CI, 9-29), and 28.7% (95% CI, 21-40) versus 10.0% (95% CI, 5-22) for patients who did and did not receive PCI, respectively (Fig. 3). The median survival time was 21.7 months (95% CI, 17-36.8) for PCI group and 11.2 months for the patients who did not receive PCI (95% CI, 8.9-14.1). Adjusted for patient age in the multivariate analysis, the patients who did not receive PCI had a significantly higher risk of death versus patients who received PCI with HR 2.0 (95% CI, 1.4 to 2.8; P = .0001).
PCI has become an important part of the management of SCLC. Initially, data from randomized trials demonstrated that PCI significantly reduced the incidence of brain metastasis but did not significantly improve OS.8-10 Data from meta-analyses that included randomized trials of patients who did and did not receive PCI demonstrated a significant improvement3, 4 in OS, 5.4% at 3 years when patients with LS-SCLC in complete remission received PCI.4 The optimal fractionation of PCI in LS-SCLC appears to be 2500 cGy in 10 fractions, as this was associated with improved OS compared with alternative PCI schedules.11
In SCLC patients with a brain-only relapse, merely 50% of patients respond to therapeutic cranial irradiation, and survival after therapeutic cranial irradiation in patients with SCLC is poor.12 In patients with a complete response to therapeutic cranial irradiation, only 37% were free of progressive brain tumors at 1 year. The median survival in patients treated with therapeutic cranial irradiation who previously received PCI is less than 2 months.13
The toxicity of PCI is a concern and has been the focus of several clinical trials assessing neurocognitive and quality of life (QOL) endpoints.8, 9, 14-17 Neuropsychological deficits are common in long-term survivors of SCLC. Deficits were found in 86% of patients who survived 6 to 13 years from treatment.14 However, the etiology of these deficits is controversial and the potential for neuropsychological toxicity from PCI remains a concern. Retrospective studies in the 1980s raised concerns over the potential neuropsychological toxicity of PCI such as ataxia, dementia, and cognitive decline.14, 15 These studies were retrospective and in many instances, PCI was delivered in large fraction sizes or concurrently with chemotherapy, unlike more modern PCI.14-16 The issue of neuropsychological toxicity with PCI was examined prospectively in 2 randomized studies;in both trials, PCI was not associated with significant neuropsychological toxicity.8, 9 Neuropsychological function was measured at the time of randomization and in follow-up. Only 41% of patients had no neuropsychological abnormality at randomization before receiving PCI.9 The high level of neurological abnormalities at randomization may have resulted from prior chemotherapy, paraneoplastic syndromes, or micrometastases in the brain. Another prospective study found that 97% of patients with SCLC had organic brain dysfunction before PCI; in these patients, PCI was not associated with a significant decline in neuropsychological function.17
Studies have assessed the use of radiotherapy in lung cancer patients and reached similar conclusions about underutilization. Delaney and colleagues found a significant discrepancy between an evidence-based recommended optimum rate of radiotherapy use in lung cancer and the actual utilization rate. Their model found that 76% of lung cancer patients should receive radiotherapy at some point in their clinical care; however, only 20% to 49% of patients do, depending on the geographic region they were treated in. Their analysis found that of all LS-SCLC patients, 94% should receive radiotherapy, whereas only 50% received radiation treatment.18 Many factors contribute to the use of radiotherapy in lung cancer in different countries, such as the type of healthcare system, access to care, availability of radiotherapy technology, and professional and patient preferences.18-20 The discrepancy between the evidence-based appropriate use of radiotherapy and actual use is not limited to lung cancer treatment but is also seen in sites such as breast cancer and gastrointestinal tract cancers.20, 21 Our review demonstrates that in a clinical setting, where radiotherapy is available to all eligible patients, patient preference and concerns about toxicity are significant factors limiting the use of radiotherapy.
Our analysis found that patients older than age 65 years were less likely to receive PCI. In other studies, older patients were found to have a higher frequency of omitted doses of chemotherapy22 and lower rates of PCI use.23 However, a retrospective review of 2 Canadian randomized trials reported there was no significant difference in the tolerance or efficacy of thoracic radiotherapy in elderly patients with LS-SCLC when compared with younger patients.24 This population of patients may require extra attention and education to ensure adequate treatment is provided.
This study is limited by the single-institutional, retrospective nature of the review. In our study, the group receiving PCI had a 24.1% and 18.7% better OS at 3 and 5 years, respectively, compared with those who did not receive PCI. This is a greater difference than that previously reported in meta-analysis data4 and is likely related to the nonrandomized nature of this dataset. The authors attempted to limit the bias inherent in comparing 2 nonrandomized groups. However, there are likely factors beyond the presence or absence of PCI treatment that contribute to the improved OS seen in the PCI group. For example, 12 patients were not offered PCI because they had a partial response to chemoradiation. Response data were not available for all patients in this review, and a difference in the number of patients with complete versus partial response in either group may contribute to the OS difference. In addition, 15 patients were lost to follow-up. As all of these patients did not receive PCI, we opted to record them as not developing brain metastases, thus, generating the most conservative assessment of difference between brain FFS in the PCI and no PCI groups. By comparison, we were able to obtain date of death details for these patients from our central provincial cancer registry, thus limiting the chance of bias in the OS analysis. As in all retrospective studies, it is difficult to control selection biases for treatment groups.
At our institution, only 61% of patients with LS-SCLC received PCI during the study period. There was no significant increase in PCI utilization in recent years, despite evidence supporting its role in management of LS-SCLC.3, 4 The most common reason for patients not to receive PCI was patient refusal, largely due to concerns about the potential toxicity of PCI. This analysis is limited by the retrospective nature of the paper.
Further prospective work is required to determine the specific concerns of patients regarding PCI and what further information is required to ameliorate these concerns. These data represent a single-institutional experience. These results should be correlated with other institutions to form a representative picture of LS-SCLC patients' concerns about PCI. However, to our knowledge, this is the first study to publish PCI utilization rates.
Not all eligible LS-SCLC patients are receiving PCI, despite its ability to improve the incidence of brain metastases and OS. The main reason for not receiving PCI is patient refusal mainly due to concerns over toxicity. PCI utilization has not significantly increased in recent years. Thorough exploration of patient concerns about PCI may increase use.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 8Prophylactic cranial irradiation is indicated following complete response to induction therapy in small cell lung cancer: results of a multicentre randomised trial. United Kingdom Coordinating Committee for Cancer Research (UKCCCR) and the European Organization for Research and Treatment of Cancer (EORTC). Eur J Cancer. 1997; 33: 1752-1758., , , et al.
- 11Standard-dose versus higher-dose prophylactic cranial irradiation (PCI) in patients with limited-stage small-cell lung cancer in complete remission after chemotherapy and thoracic radiotherapy (PCI 99-01, EORTC 22003-08004, RTOG 0212, and IFCT 99-01): a randomised clinical trial. Lancet Oncol. 2009; 10: 467-474., , , et al.