Multimodality Therapy for Limited-Stage Small Cell Lung Cancer: Combining Chemotherapy and Radiotherapy
Sara Erridge
Nevin Murray
Andrew T. Turrisi III
INTRODUCTION
The disappearance of a large malignant mass is a dramatic phenomenon in medicine. Experienced lung cancer oncologists agree that it is easy to make the chest radiograph or computed tomography (CT) scan of a patient with locally advanced small cell lung cancer (SCLC) look much better with chemotherapy and thoracic irradiation. However, it is definitely more challenging to eliminate the last clonogenic cell in the neoplasm to prevent the development of an incurable relapse. The treatment of limited-stage SCLC requires us to combine chemotherapy and radiotherapy in the most effective way to increase the proportion of long-term survivors.
The incidence of SCLC, which accounts for about 12% to 15% of all lung cancers, is decreasing in developed countries most likely because of changes in the composition of cigarettes and patterns of tobacco consumption.1 The declining incidence of SCLC parallels a stalling in the pace of research in this cancer as reflected by the number of abstracts submitted to the American Society of Clinical Oncology annual meetings and the International Association for the Study of Lung Cancer world conferences.2 In contrast, the amount of research for non-small cell lung cancer (NSCLC) has increased dramatically. The slow pace of SCLC investigation is unfortunate and puzzling because the proportion of estimated deaths from this disease is about 4% of all cancer mortality and similar to ovarian cancer, leukemia, and non-Hodgkin lymphoma.3
STAGING
SCLC may be staged by either the Veterans Administration Lung Study Group (VASLG)4 system or the tumor, node, metastasis (TNM) classification.5 The VASLG system categorizes the stage of SCLC as either limited (LSCLC) or extensive (ESCLC). This system has persisted for SCLC because of its simplicity, reliable prognostic value, and practical utility.6,7,8 LSCLC is defined as tumor confined to one hemithorax and the regional lymph nodes, whereas extensive-stage disease (ESCLC) is defined as disease beyond these bounds. The original definition of limited disease was a tumor volume that could be encompassed by a “reasonable” radiotherapy treatment plan. Because long-term survival is uncommon (7% to 9%) when chemotherapy alone is used to treat LSCLC,9,10 the reasonable radiotherapy port rule continues to be of practical importance in the design of combined modality therapy.
Although the term reasonable lacks precision, the adoption of this criterion internationally has broad acceptance in operational definition of limited disease. There continues to be support for moving to the TNM system, but this has not been widely adopted. Using simple staging techniques, the University of Toronto Lung Oncology Group8 identified a subgroup of patients with “very limited” SCLC without mediastinal node involvement who had a significantly better prognosis and a 5-year survival of 18% with sequential chemotherapy followed by thoracic irradiation used between 1976 and 1985. The 5-year survival for patients with evidence of involved mediastinal nodes was 6%. Only 2% survived 5 years when there was pneumonic consolidation, atelectasis, pleural effusion, or involved supraclavicular nodes. Such distinctions may be addressed by application of the TNM staging system. Until 2007, the TNM system had been reported only in small surgical series. At the 12th World Conference on Lung Cancer, the International Association for the Study of Lung Cancer Staging Committee presented an analysis of 8088 patients.5 Survival outcome was superior in patients without mediastinal or supraclavicular nodal involvement. Patients with pleural effusion regardless of the cytology have an intermediate prognosis between limited and extensive-stage disease. The median survival of stage IIIA/IIIB (12.1/11.1 months) and the 5-year survival (10% to 12%) reported in this data set are lower than what has been reported for state-of-the-art treatment described in this chapter. It remains to be seen whether adoption of the TNM system will improve the investigation and management of LSCLC. New and ongoing studies of SCLC treatment should report both the VASLG as well as the TNM system.
A legitimate question in a disease, such as SCLC with early widespread dissemination, is the clinical necessity of multiple staging investigations. Because chemotherapy is recommended for all fit patients, to expedite treatment, most clinicians perform a minimum number of staging procedures, rather than the wider range recommended in clinical trials. This approach might be appropriate if the management were initial treatment with multiple chemotherapy cycles followed by consolidative thoracic irradiation. On the other hand, if patients with true LSCLC are better treated by concurrent and early chemoradiotherapy, careful staging is imperative to properly categorize intent (curative vs. palliative) and determine the treatment program.2 LSCLC patients have curative potential, which justifies the complexity and toxicity of integrated thoracic irradiation and combination chemotherapy. Without appropriate staging, patients with undetected widespread disease will be subjected to unnecessary toxicity and unrealistic hope. Despite modern technology, the stage of some patients will be equivocal, so clinicians must use their judgment.
Complete evaluation of a patient with newly diagnosed SCLC consists of a history and physical examination, pathology confirmation, CT of the chest and abdomen to include the whole liver and adrenal glands, bone scan, and brain CT with contrast or MRI examination. Additionally, a complete blood count (CBC), electrolytes, blood urea nitrogen (BUN), creatinine, albumin, and liver function tests should be performed at baseline. Bone marrow biopsy has a low yield11 if other tests are negative, but may be considered in patients with leukoerythroblastic features, low platelet count or a very high lactate dehydrogenase. After such a workup, the proportion of patients with LSCLC is approximately 40%.1
The utility of positron emission tomography (PET) in SCLC has been reported in several small studies12,13,14 that have mainly used PET rather than PET/CT technology. The evidence suggests that PET added to conventional staging improves the sensitivity in detecting extracranial disease in 10% to 15% of cases, but the confidence intervals on the estimates of staging accuracy are wide because the studies are small. The information required to be confident that PET results should be used to guide therapy for LSCLC will not be available in the foreseeable future. It would be informative if a large cohort of patients with LSCLC as determined by conventional procedures also underwent a PET scan. This cohort should be treated with state-of-the-art combined modality therapy regardless of the PET result. The crucial information would not be that those with extensive SCLC according to the PET scan result have a worse outcome; this would be expected.10 The important outcome would be to demonstrate that patients with ESCLC by PET scan, but LSCLC by conventional procedures were not curable by standard treatment. It would be inappropriate for a PET scan result to deny patients with potentially curable SCLC a combined modality protocol, even if the chance for cure was decreased. On the other hand, it would be useful to know if patients with ESCLC detected by PET scan alone had an equivalent chance of long-term survival as those defined by conventional staging procedures. Whether PET imaging is useful in guiding thoracic radiotherapy planning should be evaluated in prospective clinical trials. Until this information is available, the routine use of PET scanning to guide therapy for SCLC cannot be recommended.
Although the limited-versus extensive-stage system was created for practical purposes, it also suggests important biological and clinical characteristics of the disease. When SCLC is overtly metastatic (ESCLC), an underlying biological aggressiveness in the tumor may exist that transcends the importance of the simple physical distribution of cancer cells within the body. Potentially curable LSCLC may fundamentally differ from incurable ESCLC; however, the categorical boundary may be fluid. Obliteration of the stem cells at the root cause of this disease justify early aggressive combined local management. Cure evaporates once resistant to chemotherapy stem cells establish themselves outside the primary site. Time may be of the essence when attempting to cure SCLC.
Therapeutic endeavors address concentrations of locoregional bulk of disease in LSCLC, and a subclinical metastatic population in widely distributed sites. Normal healthy tissues vary in their tolerance of therapeutic interventions, just as tumor populations vary in their susceptibility to anticancer agents. These clinical considerations and the biological factors underlying them have led directly to modern concepts of multiple-modality therapy for this disease. Although a small fraction of SCLC patients will benefit from surgery when they present with a solitary pulmonary lesion without involvement of the mediastinum, integrated chemotherapy with radiotherapy to the chest and brain offers the most realistic chance for symptom abatement, extension of median survival, and longterm survival.
THE EVOLUTION OF COMBINED MODALITY THERAPY
The legacy of treatment for LSCLC can be described by a series of treatment paradigms.15 The sequence of development of these paradigms was determined by incremental steps grounded in controlled clinical trials of crucial issues of therapy and consensus by lung cancer investigators. They provide a useful insight into how our understanding of cancer biology and treatments has evolved over the last 3 decades.
Paradigm 1: Surgery as Standard Treatment World Wars I and II profoundly influenced on the upsurge of lung cancer. The availability of tobacco products and permissive attitude toward tobacco consumption during both wars were associated with a large increase in smoking among servicemen. In addition, the prevalence of smoking among women increased greatly during the 1940s.
Surgical treatment of penetrating wounds to the chest during World War II produced major advances in thoracic surgery, and the cadre of expert chest surgeons trained during the war had to contend with the increasing numbers of lung cancers among veterans in the 1950s and 1960s.16,17 In the
immediate post–World War II era, surgery was the only truly effective treatment for patients with all types of lung cancer, including SCLC.
immediate post–World War II era, surgery was the only truly effective treatment for patients with all types of lung cancer, including SCLC.
Paradigm 2: Thoracic Irradiation Better than Surgery The epithelial origin of SCLC was described in 1926,18 and the separation of this virulent pathologic subtype on morphologic grounds was established in 1959.19 From the earliest reports,20 impressive regressions of SCLC induced by radiotherapy suggested an integral role of this modality in the definitive management of this disease.
In a trial conducted in the 1960s, the median survival of surgically unresectable LSCLC patients randomized to supportive care alone was 12 weeks.21 Another randomized trial comparing surgery alone with thoracic irradiation alone for patients with SCLC was carried out by the Medical Research Council in the United Kingdom.22,23 Eligibility criteria included (a) SCLC on bronchial biopsy; (b) no evidence of extrathoracic metastasis; (c) the tumor was regarded as operable on clinical examination and chest radiograph; (d) the patient was considered fit enough for resection; and (e) the patient was considered fit enough for radical radiotherapy. Although this study was conducted in an era before the availability of modern staging techniques, the intrathoracic extent of tumor was probably less than in most contemporary LSCLC trials. Of the 144 patients included in the main analysis, 71 were allocated to surgery and 73 to radical megavoltage radiotherapy. A complete resection of all visible growth was performed in 48% of the surgical group, all of whom had a pneumonectomy. A total of 34% were unresectable, and 18% had no operation because of preoperative deterioration or refusal of surgery. In the radiotherapy group, 85% had radical thoracic irradiation, 11% had palliative radiotherapy, and 4% had no radiotherapy because of deterioration or refusal. The median survival was 28.5 weeks for surgery and 43 weeks for radiotherapy (p = 0.04). Five-year survival was 1% for the surgical arm (the sole survivor refused surgery and was given radiotherapy) and 4% for radiotherapy. Outcome for both groups was poor, but treatment feasibility, toxicity, and survival all favored thoracic irradiation. The standard of treatment for LSCLC shifted from surgery to thoracic irradiation. The main aim was to give patients relief of local symptoms until their death from metastatic disease.
Paradigm 3: Thoracic Irradiation with Adjuvant Chemotherapy The systemic nature of SCLC was emphasized by the rapid tempo of systemic relapse and low probability of long-term survival in patients with apparently localized disease given definitive local therapy alone. In a classic study24 of 19 patients undergoing potentially curative surgical resection who died of noncancer-related causes within 30 days of surgery, 13 were found to have persistent disease at autopsy. Moreover, distant metastases were present in 12 of the 13 cases. Although not all patients with LSCLC have subclinical metastatic disease, the actual proportion is so high that we assume that there are metastases and treat accordingly. The success of chemotherapy for leukemia and lymphoma was in its infancy, but already the vision of cure by systemic treatment stimulated clinical research into using such an approach for SCLC.
A major step in the systemic treatment of SCLC was reported in 1969. This randomized study compared alkylating agents at several dose schedules with an inert compound in about 2000 patients with lung cancer at a group of Veterans Administration hospitals.4 The antitumor effects of chemotherapy were analyzed according to cell type, and improvement in survival was the sole criterion of drug activity. The 4-month median survival for patients with SCLC treated with high intermittent doses of cyclophosphamide compared with 1.5 months for patients given placebo (p = 0.0005). Documentation of a survival improvement with chemotherapy for patients with lung cancer was a notable development in cancer medicine, and cyclophosphamide became the cornerstone in SCLC chemotherapy regimens for decades. The credibility of cyclophosphamide efficacy in the treatment of SCLC was augmented by randomized trials that showed prolonged survival for that agent as adjuvant chemotherapy compared with no further treatment after surgical resection.17 Curiously, the lack of survival benefit for cyclophosphamide and other alkylating agents in the treatment of NSCLC in these and other lung cancer trials17,25,26,27,28,29,30 did not prevent them from being incorporated into combination chemotherapy regimens for this disease over the next 20 years.
The perceptive observations of Watson and Berg20 suggested thoracic irradiation coupled with chemotherapy as a model for treatment of SCLC. This hypothesis was first tested in a randomized trial by Bergsagel et al.31 from Toronto. Patients with nonresectable lung cancer confined to the central area of the thorax were randomly assigned treatment with radiotherapy to the primary lesion and mediastinum or radiotherapy plus two schedules of intermittent intravenous cyclophosphamide. One third of 123 patients in the study had SCLC and both progression-free survival (29 vs. 16 weeks) and overall survival (42 vs. 21 weeks) were significantly superior for patients receiving combined modality therapy.
Two other randomized trials32,33 showed significantly superior survival for thoracic irradiation and adjuvant chemotherapy when compared with radiotherapy alone. The median survival for patients with LSCLC treated by radiotherapy alone was consistently about 5 to 6 months. Two additional randomized controlled trials34,35 showed some advantage for combined modality therapy, but the survival differences between the groups were not statistically significant, probably because many patients in the radiotherapy arms received chemotherapy at the time of disease progression.
The paradox that a local modality had a role in a disease dominated by systemic spread had been established. The issues of integration such as what drugs, what timing, what volume to irradiate, and what dose were not clear enough to formulate and endure today.
Paradigm 4: Combination Chemotherapy with Adjuvant Thoracic Irradiation The success of combination chemotherapy in leukemia and lymphoma and the recognition of SCLC as a type of lung cancer with marked
chemosensitivity spurred investigation of multidrug regimens. The first study of combination chemotherapy for lung cancer was published in 1972 by Hansen et al.,36 using a regimen that included cyclophosphamide, methotrexate, dactinomycin, and vincristine. All eight patients with SCLC subtype responded; combination chemotherapy for SCLC was off to a promising start. High response rates for cyclophosphamide and vincristine were reported by Eagan et al.37 and Holoye and Samuels.38 By combining cyclophosphamide, doxorubicin, vincristine (CAV), and bleomycin, Einhorn et al.39 produced not only high response rates in SCLC, but complete responses were seen in 20% of cases. Bleomycin was discarded because of pulmonary toxicity, especially when combined with thoracic irradiation,40 and the CAV regimen persists to this day as a standard regimen for SCLC. A cardinal feature of drug selection was individual agent activity in SCLC, nonoverlapping toxicity of each of the agents in the combination, and the holy grail of synergy.
chemosensitivity spurred investigation of multidrug regimens. The first study of combination chemotherapy for lung cancer was published in 1972 by Hansen et al.,36 using a regimen that included cyclophosphamide, methotrexate, dactinomycin, and vincristine. All eight patients with SCLC subtype responded; combination chemotherapy for SCLC was off to a promising start. High response rates for cyclophosphamide and vincristine were reported by Eagan et al.37 and Holoye and Samuels.38 By combining cyclophosphamide, doxorubicin, vincristine (CAV), and bleomycin, Einhorn et al.39 produced not only high response rates in SCLC, but complete responses were seen in 20% of cases. Bleomycin was discarded because of pulmonary toxicity, especially when combined with thoracic irradiation,40 and the CAV regimen persists to this day as a standard regimen for SCLC. A cardinal feature of drug selection was individual agent activity in SCLC, nonoverlapping toxicity of each of the agents in the combination, and the holy grail of synergy.
By using the principle of combination chemotherapy41 and incorporating new, more active agents, the search for a regimen with a high complete response rate for SCLC was intense during the 1970s. Combination chemotherapy was shown to be better than single-agent chemotherapy in three early randomized trials,42,43,44 but a superior regimen did not emerge. Nevertheless, in the early 1970s, a change in philosophy occurred; with the observed high response rates, multiagent chemotherapy became the primary therapy in LSCLC and thoracic irradiation was positioned as an adjuvant or “consolidative” treatment after initial numbers of cycles of systemic therapy.45
Treatment regimens giving aggressive combination chemotherapy without thoracic irradiation46 appeared to yield survival results similar to combined modality therapy. Median survival was in the range of 12 to 15 months, and projected long-term survival was usually in the range of 10%, whether radiotherapy was administered or not. However, aggressive combined modality therapy was also associated with more toxicity, and the selection of only patients fit enough to receive this more demanding treatment in nonrandomized reports may have biased results against chemotherapy alone. Many investigators began to speculate that radiotherapy might not be necessary at all in LSCLC.
This debate persisted throughout the 1980s,45,47 and many randomized trials were performed in an attempt to settle this vexing issue.48,49,50 It was not until 1992 that two metaanalyses9,10 demonstrated a modest improvement in survival rates in those patients given thoracic radiotherapy in addition to chemotherapy. The survival benefit becomes evident at about 15 months after the start of treatment and persists beyond 5 years. At 3 years, 8.9% of the chemotherapy-only group was alive compared with 14.3% of the combined modality group. The analysis of local control showed a 2-year local failure rate of 23% for irradiated patients versus 48% for nonirradiated patients (p = 0.0001). These benefits were obtained at the cost of an increase in treatment-related deaths of 1%. However, none of the trials in these metaanalyses employed initial cisplatin etoposide chemotherapy, which is now the acknowledged international standard.
The addition of thoracic irradiation to chemotherapy for LSCLC has, for several theoretical reasons, the potential to improve outcomes. Its mechanism of action is different from chemotherapeutic agents so the potential exists for additive or even synergistic damage to the tumor. This capacity to eradicate the most populous, and hence most dangerous concentration of tumor cells, improves the probability of controlling local disease that may evolve more resistant progeny and metastasize systemically
Paradigm 5: Integrated Early Concurrent Chemotherapy and Radiotherapy The strategy to destroy as many cancer cells as possible in the shortest period of time using early concurrent chemoradiation has several theoretical advantages.
Decreased Probability of Metastatic Events Experimental work by Hill et al.51 indicate that tumor cells mutate spontaneously and randomly to acquire metastatic potential. Moreover, once tumors reach a critical size or volume, metastatic phenotypes are generated “explosively,” so the cumulative probability of the existence of metastases and the number of metastases increases in proportion to elapsed time. The best way to decrease metastatic events is to quickly eliminate as much tumor as possible.
Lower Probability of Chemotherapy Resistance A large body of experimental and clinical data exists that support the observation that variability exists for chemosensitivity within tumor cell populations.52 Moreover, tumor cells display a capacity to be resistant to many drugs concurrently.53 The biologic basis of this evolution of resistance originates during tumor growth from mutations in the cancer genome.54 The development of resistant mutants is a random process, and the probability of their appearance increases with time in proportion to the total number of cell divisions the neoplastic burden has undergone. The best way to minimize the probability of chemotherapy resistance is to eliminate as many cancer cells as possible in the shortest time.
Lower Probability of Resistance to Radiotherapy The probability of mutation to intrinsic radiotherapy resistance55 or acquired radioresistance as a consequence of enhanced DNA repair efficiency secondary to previous chemotherapy56 should be minimized by the early deployment of both modalities.
Diminished Accelerated Repopulation Accelerated repopulation of tumors undergoing radiotherapy has been proposed57 to explain the clinical observation that regimens with extended duration of therapy often require increased radiation dosages to achieve an isoeffective result.58 Accelerated tumor growth has been reported after surgery59 and chemotherapy60 in animal models. Accelerated repopulation will decrease local control, and also, the increased mitotic activity within
a larger residual tumor may result in an increased probability of metastatic events and the development drug and radiation resistance. Rapid destruction of tumor by early integration of chemoradiation should minimize the amount of tumor capable of repopulation.
a larger residual tumor may result in an increased probability of metastatic events and the development drug and radiation resistance. Rapid destruction of tumor by early integration of chemoradiation should minimize the amount of tumor capable of repopulation.
Although theoretically attractive, the practical implementation of early chemoradiation initially proved challenging. In 1976, investigators from the Radiation Oncology Branch of the National Cancer Institute in Bethesda, Maryland performed an exploratory study that tested the limits of toxicity.61,62,63,64,65 Although the protocol was influenced by the successful model of combined modality therapy for childhood lymphocytic leukemia,66 the vigor of therapy was unprecedented in solid tumor protocols. It involved initial simultaneous irradiation to the brain, primary tumor, and mediastinum and aggressive concurrent chemotherapy (cyclophosphamide 1.5 g/m2, doxorubicin 40 mg/m2, and vincristine 2 mg). The drugs were repeated as soon as the leukocytes increased to 3.5 × 109/L. All therapy was complete in 3 to 4 months.
The toxicity of this regimen was formidable. Radiation pneumonitis occurred in 38%, and a combination of pneumonitis and neutropenic sepsis was fatal in 24%. Severe esophagitis requiring nasogastric or parenteral nutrition occurred in 14%, and permanent strictures were observed.62 Additionally, a previously undescribed neurologic syndrome of somnolence, poor attention span, recent memory loss, and action tremor was seen. The symptoms became evident within 2 to 4 months of starting treatment and were reversible within 4 months of onset.
The first reported survival results for patients with LSCLC treated in this manner were spectacular with 100% complete remissions and projected 80% long-term survival.61 With longer follow-up, the survival rates dwindled and this trial was criticized for generation of false optimism by preliminary data reporting and unacceptably severe toxicity. Nevertheless, the mature results, which demonstrated an 80% complete remission rate and 25% survival at 4 years, were provocative and unprecedented.65 Although the response result was noteworthy, 13% died without evidence of tumor at autopsy; it is speculative whether the survival rate could have been higher had modern supportive care been available or whether the toxicity of the approach was unacceptable. In an analysis of treatment factors contributing to long-term survival, it was concluded that concurrent chemotherapy and radiotherapy achieved better local tumor control than sequential therapy.65 These data infer that if the duration of concurrent therapy is prolonged longer than 3 weeks, the enhanced survival and local tumor effect were lost in a flood of treatment-induced toxicity.55
It was evident that the combined chemoradiation model could not progress unless chemotherapy and radiotherapy could be compatible and integrated in a manner such that extreme toxicities did not compromise delivery of either modality. Sequential regimens giving thoracic irradiation during a gap in chemotherapy67 (the “sandwich” technique) were less toxic but efficacy was not improved, possibly because the interruption of the cadence of chemotherapy allowed tumor regrowth in nonirradiated sites. The results of a phase II study of a less aggressive CAV regimen68 with concurrent thoracic irradiation led to a large randomized trial comparing simultaneous chemoradiation with CAV chemotherapy alone. However, a significant survival benefit was not observed69 and excessive toxicity impaired drug delivery. Another controlled trial of CAV alone versus a split course of thoracic irradiation delivered in three phases between CAV pulses demonstrated a survival benefit for combined modality therapy.50 However, gastrointestinal and hematologic toxicity continued to be problematic. The severity of morbidity to normal tissue from the interaction of doxorubicin and radiotherapy has limited the utility of these approaches. Concurrent thoracic irradiation with chemotherapy containing cyclophosphamide, methotrexate, and lomustine was superior to chemotherapy given alone,48 with higher rates of remission and prolonged survival in patients with LSCLC, but the benefit was offset by unacceptable pulmonary toxicity. Similarly, concurrent cyclophosphamide, etoposide, and vincristine and radiotherapy have demonstrated superiority to chemotherapy alone, but unacceptable chemotherapy attenuation from myelosuppression was associated with combined therapy.49