Michael F. Gensheimer and Billy W. Loo Jr.

INTRODUCTION

Small cell lung cancer (SCLC) is a high-grade neuroendocrine carcinoma of the lung that is prone to early metastatic spread and often responds well to chemotherapy and radiation therapy (RT), but tends to recur after treatment. Around 40% of patients present with limited-stage disease (LS-SCLC), which is potentially curable with RT, chemotherapy, and/or surgery. The remaining 60% of patients present with typically incurable, extensive-stage disease (ES-SCLC) (1). Survival after treatment for LS-SCLC is poor due to frequent disease recurrence. Most patients with LS-SCLC have stage III disease with mediastinal lymph node involvement and a median survival of 12 to 14 months (1). In fit patients, the goal of treatment for LS-SCLC is cure, but treatment can also palliate symptoms.

Initial Evaluation

At a minimum, patients should undergo CT of the chest and abdomen with IV contrast, laboratory tests including complete blood count and serum chemistries, and brain MRI. Fluorodeoxyglucose-PET (FDG-PET)/CT scan is also recommended as it is more sensitive than CT alone, particularly for the detection of distant metastases. In a collection of prospective and retrospective studies, PET scan findings altered initial management in 28% of patients, either due to stage shift or a change in the RT field (2). If conventional imaging or biopsy shows clear evidence of distant metastatic disease, then PET/CT may be omitted. Patients with clinically lymph node-negative disease who are being considered for surgical resection should undergo invasive mediastinal staging (e.g., mediastinoscopy or endobronchial ultrasound-guided biopsy) prior to tumor resection since identification of mediastinal lymph node involvement would change local treatment from surgery to RT.

Stage Classification

Less than 5% of patients with SCLC present with localized disease that is amenable to surgical resection. Most patients have locally advanced or metastatic disease and will be treated most appropriately with either chemotherapy plus RT or chemotherapy alone. Determining which patients are candidates for definitive RT was the rationale for the influential 1973 Veteran’s Administration Lung Group (VALG) staging system (3). By this system, LS-SCLC was defined as disease in the ipsilateral hemithorax that can be safely encompassed within a tolerable radiation portal. All other patients were defined as having ES-SCLC. This two-stage system is of historical importance for understanding the SCLC literature, including the many clinical trials that have used the VALG system or minor variations of it. Problems with this system include the lumping together of patients with varied prognoses into the LS-SCLC designation, and the subjectivity of defining a tolerable radiation portal.

More recently, the American Joint Committee on Cancer (AJCC) seventh edition adopted a uniform tumor/node/metastasis (TNM) staging system for both non–small cell and SCLC as it has been shown to be prognostic in both diseases (1). The TNM system should be used because of its more precise therapeutic implications and to permit stage-based analysis of clinical outcomes going forward. By combining the functional nature of the VALG two-stage system with the granularity of the AJCC TNM system, the National Comprehensive Cancer Network (NCCN) Guideline for SCLC defines limited-stage as AJCC stage I to III (T any, N any, M0) that can be safely treated with definitive RT doses, excluding T3-4 due to multiple lung nodules that are too extensive or have tumor/nodal volume that is too large to be encompassed in a tolerable RT plan, and extensive-stage as AJCC stage IV (T any, N any, M1a/b) or T3-4 due to multiple lung nodules that are too extensive or have tumor/nodal volume that is too large to be encompassed in a tolerable RT plan. Figure 10.1 shows PET scans from patients with AJCC stage IIIA/LS-SCLC and AJCC stage IV/ES-SCLC.

TREATMENT OF LS-SCLC

Standard treatment for LS-SCLC involves four to six cycles of chemotherapy with definitive thoracic RT given early during the chemotherapy course. Patients with radiographic response to treatment and without progression after completion of chemotherapy and RT should be considered for prophylactic cranial irradiation (PCI). Rare patients with very limited extent of disease (T1-2, N0, and M0) can be considered for surgery followed by adjuvant chemotherapy instead of concurrent chemoradiotherapy.

Figure 10.1 PET Images of Two Patients With Small Cell Lung Cancer: (A) AJCC Stage IIIA/VALG Limited-Stage Confined to the Right Lung and Ipsilateral Hilar/Mediastinal Lymph Nodes; (B) AJCC Stage IV/VALG Extensive-Stage With Bone and Extra-Thoracic Lymph Node Metastases

Chemotherapy

SCLC is highly sensitive to chemotherapy and early chemotherapy drugs, such as alkylating agents, showed activity in this disease. Due to improved outcomes with combination regimens over single agents, cyclophosphamide, doxorubicin, and vincristine (CAV) became a standard regimen for SCLC. More recently, the combination of etoposide plus cisplatin (EP) showed less toxicity and similar to improved efficacy when compared to alkylator/anthracycline-based regimens in several randomized trials. A Norwegian Lung Cancer Study Group trial comparing EP to cyclophosphamide, epirubicin, and vincristine included 436 patients with LS-and ES-SCLC (4). Patients with LS-SCLC received thoracic RT with cycle 3 of chemotherapy. Survival was improved in the EP arm (median, 10.2 vs. 7.8 months, P = .004), and the survival benefit appeared to be confined to patients with limited-stage disease (median, 14.5 vs. 9.7 months), perhaps because EP was better tolerated in combination with RT. For example, EP avoids synergistic toxicity, particularly cardiac toxicity, between anthracyclines and RT (5).

Cisplatin Versus Carboplatin

Many patients with SCLC are not good candidates for cisplatin, due to comorbidities such as preexisting renal impairment, neuropathy, or hearing loss. Several trials have studied the combination of etoposide plus carboplatin (EC) as an alternative to EP. An early trial from Greece compared six cycles of EC to six cycles of EP in 147 patients with either LS-SCLC or ES-SCLC (6). Patients with LS-SCLC received thoracic RT with cycle 3. Median survival was not different between the study arms, at around 12 months in each. Severe toxicity was more frequent with EP, including leukopenia, neutropenic fever, nausea, vomiting, and neurotoxicity. A subsequent meta-analysis of a trial that compared cisplatin-based versus carboplatin-based regimens in SCLC also found equivalent survival and response rates (7). Hematologic toxicity was more common with carboplatin, but nonhematologic toxicity was greater with cisplatin. In summary, it is reasonable to substitute carboplatin for cisplatin in combination with etoposide, especially in patients with cisplatin-specific toxicity concerns (8).

Alternate Regimens

Trials evaluating alternative chemotherapy regimens for LS-SCLC have not shown superiority to EP. In a recent Japanese trial, all patients received one cycle of EP concurrently with RT prior to randomization to three more cycles of EP or three cycles of cisplatin plus irinotecan (9). Overall survival was not improved with irinotecan and toxicity was similar between the two arms. Targeted agents and immunotherapy are currently being evaluated in patients with ES-SCLC, and if found to provide survival benefit they would eventually be assessed in LS-SCLC. For instance, in a phase I/II trial in patients with ES-SCLC, the immune checkpoint inhibitor nivolumab, given with or without the CTLA4 inhibitor ipilimumab, induced durable responses lasting up to 11 months (10).

Elderly and Poor Performance Status Patients

Standard therapy for LS-SCLC with platinum-based chemotherapy plus concurrent RT is toxic. For instance, in a recent trial of fit patients (≤70 years old, ECOG performance status 0–1), four cycles of EP plus RT resulted in a 95% rate of neutropenia and a high rate of neutropenic fever (25% during concurrent chemoradiotherapy plus an additional 16% during consolidation chemotherapy) (9). Therefore, it would be helpful to deintensify treatment for LS-SCLC, especially for older patients or those with poor performance status.

Published studies have not shown encouraging results with treatment deintensification for SCLC. One study enrolled 300 patients who were ≤75 years old with either LS-SCLC or ES-SCLC, and randomized them to chemotherapy with cyclophosphamide, vincristine, and etoposide, given either up-front or at time of symptomatic or radiographic progression (11). Median survival was the same in the two arms (36 weeks for up-front and 32 weeks for delayed chemotherapy, P = .96). However, the delayed chemotherapy arm had inferior quality-of-life, specifically in the domains of mood, sleep, and general well-being. This suggests that there is value in preemptive palliation, or treatment given before symptoms become severe. In a second trial, patients who were ≥75 years old with any stage of SCLC, or younger patients with ES-SCLC, were randomized to a standard chemotherapy regimen (EP alternating with CAV) versus oral etoposide, which was expected to be better tolerated (12). However, progression-free survival (PFS) was inferior in the oral etoposide arm (median, 3.6 vs. 5.6 months, P = .001), with a trend toward inferior overall survival (median, 4.8 vs. 5.9 months, P = .13). Quality-of-life was also inferior in the oral etoposide arm, presumably due to symptoms of progressive cancer.

While these two trials do not provide support for delayed or deintensified treatment for LS-SCLC, patients unable to tolerate full-intensity therapy due to comorbidities or performance status were not included in most randomized trials in LS-SCLC and their treatment should be individualized. If there are concerns that an individual patient will not tolerate concurrent chemoradiotherapy, starting treatment with chemotherapy alone is a reasonable option with the addition of RT with a later cycle if the patient tolerates treatment well. Sequential chemotherapy followed by definitive RT could also be considered. For patients with very poor performance status, palliative RT without chemotherapy, or supportive care without active anticancer treatment are appropriate options.

Thoracic RT

Clinical trials reported in the 1970s and 1980s showed that the addition of RT to chemotherapy improved local control in the chest, but some did not demonstrate a significant improvement in overall survival. Two meta-analyses, both published in 1992, helped clarify this issue by reporting an improvement in overall survival with the addition of RT (13, 14). The meta-analysis reported by Pignon et al. analyzed individual data on 2,103 patients with LS-SCLC enrolled in 13 trials that compared chemotherapy with or without RT (13). In some of the trials, RT was delivered concurrently with chemotherapy, while in others it was given after the completion of chemotherapy. The relative risk of death in the combined-therapy group was 0.86, with an increase in 3-year overall survival from 9% to 14%. Most of the benefit was seen in patients younger than 65 years old.

Timing of RT

Once concurrent chemotherapy and thoracic RT had been established as the standard-of-care, the next question to be answered was whether RT should be given early or late in the chemotherapy course. A few influential studies showed that early RT given with the first or second cycle of chemotherapy likely improves outcomes. In a trial performed by the National Cancer Institute of Canada, 308 patients all received six cycles of CAV alternating with EP chemotherapy with randomization to undergo thoracic RT (40 Gy in 15 fractions) with either the first or last cycle of EP (15). In both arms, 83% of patients completed all six cycles of chemotherapy, but more patients on the early thoracic RT arm received RT (96% vs. 87%). Overall survival was improved in the early thoracic RT arm (median, 21 vs. 16 months, P = .006), with a sustained separation of the survival curves (5 years, 20% vs. 11%). A subsequent meta-analysis of thoracic RT timing by Fried et al. analyzed 1,524 patients from seven trials, though they did not have access to individual patient data (16). Early RT was defined as beginning with cycle 1 or 2 of chemotherapy and less than 9 weeks after the start of chemotherapy. A small, but statistically significant, benefit in 2-year overall survival was seen in the early RT arm (relative risk of death 0.85, P = .03). Finally, another meta-analysis analyzed overall survival according to time from start of any treatment to the end of radiotherapy (SER), and found the best survival when SER was less than 30 days, with a decrease in 5-year survival rate of around 2% per week of extension of SER (P = .0003) (17).

In summary, thoracic RT should be started with the first or second cycle of chemotherapy, and the time from initiation of treatment to completion of RT may be a valuable predictor of treatment effectiveness. Although RT should be started with cycle 1 whenever possible, in specific situations it is acceptable to start with cycle 2 or even later. First, if a patient is symptomatic and needs to start treatment urgently, it may be best to start chemotherapy before RT to allow time for careful RT planning. Second, if disease is very bulky, it may be useful to use chemotherapy to downsize the treatment volume and reduce the risks of RT-induced toxicity, such as radiation pneumonitis. A prospective trial demonstrated that targeting RT to the postchemotherapy extent of disease does not compromise disease control (18). In this study, RT was given with cycle 3 of chemotherapy and patients were randomized to RT targeting the pre- or postchemotherapy tumor volume. The rate of isolated out-of-field local-regional recurrence was less than 3% in both arms.

RT Dose and Fractionation

The delivery of thoracic RT for SCLC has evolved over the decades as RT technology and imaging continue to improve, leading to a better understanding of appropriate doses and target volumes. A landmark randomized trial, Intergroup 0096, published by Turrisi et al. in 1999, established accelerated RT as superior to conventionally fractionated RT (19). Tolerable acceleration, or reduction of the overall duration of the treatment course, was achieved by hyperfractionation, or twice-daily administration. In the trial, 417 patients received four cycles of EP, and RT was started with the first cycle of chemotherapy. The hyperfractionated-accelerated arm received 45 Gy in twice-daily 1.5 Gy fractions over 3 weeks and the control arm received the 45 Gy in once-daily 1.8 Gy fractions over 5 weeks. Acute toxicity was higher in the accelerated arm, with 27% versus 11% of patients experiencing grade 3 to 4 esophagitis. As such, 45 Gy was considered the maximum tolerated dose when given on the accelerated schedule. However, the accelerated arm had improved overall survival (median, 23 vs. 19 months; 5-year 26% vs. 16%; P = .04). Finally, accelerated treatment also reduced the rate of local failure from 52% to 36%. The superiority of accelerated RT provides clinical evidence for the radiobiological phenomenon of accelerated repopulation of SCLC, in which rapid tumor cell growth between RT fractions later in the course diminishes the effectiveness of a given total dose of RT with prolonged treatment duration.

Despite the results of this study, accelerated RT has not been widely adopted into clinical practice in the United States. A 2003 patterns of care study showed that fewer than 10% of patients with LS-SCLC received accelerated RT, while more than 80% received once-daily treatment (20). This is likely in part due to the logistical challenge of twice-daily delivery. Furthermore, the true advantage of the hyperfractionated regimen has been questioned because the RT regimen of 45 Gy over 5 weeks used in the control arm of the Intergroup study (19), though promising in contemporaneous trials (21), is now considered suboptimal for definitive therapy, attenuating the impact of the comparison to accelerated RT. A North Central Cancer Treatment Group trial conducted at the same time as the Intergroup trial failed to show an advantage for 48 Gy of twice-daily RT compared to 50.4 Gy of once-daily RT, though a 2.5 week RT break in the twice-daily arm may have allowed accelerated repopulation (22). CALGB 8,837 demonstrated that the maximum tolerated dose when delivered in once-daily fractions of 2 Gy exceeds 70 Gy (23), a dose that is consistent with the definitive dose used for locally advanced NSCLC.

Two ongoing cooperative group trials are now comparing a higher total dose of conventional RT to a lower total dose on an accelerated schedule. The CALGB 30610/RTOG 0538 trial randomizes patients to either 45 Gy twice-daily RT over 3 weeks or 70 Gy once-daily RT over 5 weeks starting with the first or second cycle of EP. The EORTC CONVERT trial is testing a similar RT question, but all patients receive RT starting with the second cycle of chemotherapy.

RT Target Volume

Historically, due to the risk of microscopic regional lymph node involvement, patients were treated with elective nodal irradiation (ENI), which called for large fields that included the mediastinum, bilateral lung hila, and supraclavicular fossae even if these regions were not clearly involved with the cancer. There has been a steady trend toward the use of smaller fields, in an attempt to reduce RT-related toxicity, such as esophagitis and pneumonitis. This trend has converged toward involved-field irradiation (IFI), in which only areas that are clinically involved on PET or CT are targeted (Figure 10.2). Several single-institution studies suggest that IFI is associated with low rates of out-of-field regional recurrence when PET-CT is used to target involved nodes that might be missed by CT alone. In two of these studies, only 3 of 120 patients (2.5%) had isolated out-of-field regional failure (24,25). Current clinical trials, such as CALGB 30,610, have adopted the IFI approach. Of note, with reduced treatment volumes, it is likely that the maximal tolerated dose would exceed 45 Gy, even with an accelerated hyperfractionated schedule, but this issue has not yet been evaluated in a clinical trial.

RT Summary

While still undergoing evolution, expert consensus on standard practice currently recommends IFI treatment volumes using dosing regimens ranging from 45 Gy of accelerated hyperfractionated RT over 3 weeks to 60 to 70 Gy of conventionally fractionated RT over 6 to 7 weeks (8). Expected outcomes with this strategy include a complete radiographic response rate of 50%, median overall survival of around 2 years, a 5-year overall survival rate of 25%, and a local-regional failure rate of 25% to 35% (19,18). The majority of disease recurrences are at distant sites.

Stay updated, free articles. Join our Telegram channel

Join