Fax: (011) 31-43-3877055
Neurologic disorders in 432 consecutive patients with small cell lung carcinoma
Version of Record online: 15 JAN 2004
Copyright © 2004 American Cancer Society
Volume 100, Issue 4, pages 801–806, 15 February 2004
How to Cite
Seute, T., Leffers, P., ten Velde, G. P. M. and Twijnstra, A. (2004), Neurologic disorders in 432 consecutive patients with small cell lung carcinoma. Cancer, 100: 801–806. doi: 10.1002/cncr.20043
- Issue online: 3 FEB 2004
- Version of Record online: 15 JAN 2004
- Manuscript Accepted: 21 NOV 2003
- Manuscript Revised: 14 NOV 2003
- Manuscript Received: 8 SEP 2003
- lung carcinoma;
- small cell lung carcinoma (SCLC);
- neurologic complication;
- brain metastases (BM);
- leptomeningeal metastases;
- paraneoplastic syndrome
Neurologic complications are an important cause of morbidity and possibly also mortality in patients with small cell lung carcinoma (SCLC). The current study was undertaken to prospectively investigate survival and the frequency of neurologic disorders in patients with SCLC.
Between October 1980 and September 2001, 432 consecutive patients with microscopically proven SCLC were included in the current study. Patients underwent neurologic examinations on a regular basis prior to, during, and after treatment. Routine imaging of the brain (computed tomography or magnetic resonance imaging) was performed before and after systemic therapy.
A neurologic disorder was diagnosed in approximately 56% of the SCLC patients. In nearly half of the cases, the neurologic disorder already was present at the time of diagnosis. Brain metastases (BM) were diagnosed most frequently. Seventy-four patients (18%) had BM at the time of diagnosis; in 20 of these patients, the BM did not demonstrate clinical signs. Another 101 patients developed BM during follow-up. The 2-year cumulative risk of BM reached 49% for patients with limited disease (LD) and 65% for patients with extensive disease (ED). Patients with BM as the only site of disease dissemination were found to have a poorer survival compared with LD patients. The majority of the nonmetastatic disorders preceded the diagnosis of SCLC. The syndrome of inappropriate antidiuretic hormone secretion (SIADH) was diagnosed most frequently.
In this prospective study, neurologic disorders were diagnosed in greater than half of the patients with SCLC. BM were detected most frequently. Approximately 18% of the patients were found to have BM at the time of diagnosis. In approximately 33% of the cases, these BM did not cause symptoms. BM were found to have a negative effect on survival in patients with SCLC. Cancer 2004;100:801–6. © 2004 American Cancer Society.
In 2000, it was estimated that 164,100 new cases of lung carcinoma are diagnosed per year in the U.S., with 159,900 deaths from the disease reported annually.1 Of all incident cases of lung carcinoma, approximately 20–25% are reported to be small cell lung carcinoma (SCLC).2
The incidence of SCLC in the Netherlands is 15.3 (per 100,000 persons age ≥ 16 years).3 Despite new treatment regimens, survival remains poor.4 This is due in part to the neuroendocrine phenotype of SCLC, which distinguishes it from other types of lung carcinoma. These neuroendocrine properties are most likely responsible for the unique behavior of SCLC, which includes metastatic activity to the central nervous system (CNS) and the generation of paraneoplastic syndromes.5 Neurologic complications frequently are reported in patients with SCLC and are believed to contribute heavily to morbidity as well as mortality.6, 7 With a prevalence of approximately 10%, brain metastases (BM) are the most common neurologic complication diagnosed in patients with SCLC. Another 40% of patients have been reported to develop BM during follow-up.8 The probability of developing BM increases to 50–80% at 2 years from diagnosis.9
The current study was undertaken to prospectively investigate survival and the frequency of neurologic disorders in patients with SCLC. Patients were followed prospectively, and neurologic disorders were actively searched for. The current study is a continuation of a previously published study.6 Because the current study included a larger number of patients, it allowed a more detailed analysis to be performed.
MATERIALS AND METHODS
Between October 1980 and September 2001, 432 consecutive patients with microscopically and histologically proven SCLC were enrolled in the current study. Patients were diagnosed and treated at the University Hospital Maastricht (The Netherlands). Potential follow-up for each patient was at least 1 year. Patients initially were staged by a pulmonologist; physical examination, routine blood and chemistry profile, chest X-ray, computed tomography (CT) scan of the chest, and fiberoptic bronchoscopy were performed routinely. Limited disease (LD) was defined as tumor confined to one hemithorax, the mediastinum, and the ipsilateral and/or contralateral scalene and supraclavicular lymph nodes. Tumor occurring beyond these sites was defined as extensive disease (ED).
All patients were examined by a neurologist at the time of diagnosis, every 3 months during the first year, and every 6 months thereafter. Imaging of the brain was performed before and after treatment. An additional brain scan was performed when patients had survived for 12 months after the initial diagnosis of SCLC. Until 1989, CT scans were used. After 1989, CT was replaced by magnetic resonance imaging (MRI). In the case of new neurologic symptoms or signs developing, neurologic consultations and diagnostics were performed more frequently.
Initial treatment was comprised of combination chemotherapy using cyclophosphamide, doxorubicin, and etoposide. This combination was repeated at 3–4-week intervals, with a maximum of 5 courses. The actual number of courses depended on the clinical condition of the patient. Local radiotherapy occasionally was applied to the primary tumor after chemotherapy; no strict indication criteria were established for this treatment. Response after initial treatment was measured using standard criteria. Complete remission (CR) was defined as the total clinical and radiologic resolution of the disease. A reduction of at least 50% of all measurable lesions was regarded as a partial remission (PR). The remainder of responses were regarded as no response (stable disease or progressive disease).
Patients in CR were eligible for prophylactic cranial irradiation (PCI). Until 1986, both patients with LD and those with ED could opt for PCI. After 1986, only patients with LD could do so. Between 1990–1997, PCI was excluded from the treatment protocol. PCI was administered to the entire brain after the completion of the chemotherapy. The total dose was 30 grays (Gy), administered as either 3 Gy given 4 times a week or 2 Gy given 5 times a week. Clinically manifest BM were treated with whole brain radiotherapy with fractions of 3 Gy given 4 times a week, also up to 30 Gy in total dose. In the case of cerebral edema, patients received corticosteroid medication.
Patients with leptomeningeal metastases were treated with intrathecal methotrexate, systemic chemotherapy, or local radiotherapy at symptomatic sites. Therapy for epidural or intramedullar spinal metastases was comprised of radiotherapy combined with corticosteroids.
|Median age (yrs) (range)||65 (32–80)||66 (39–89)|
|Metastases (not in CNS)|
|Radiotherapy to the brain||25||65||90|
|Prophylactic cranial irradiation||35||10||45|
Survival curves were estimated using the Kaplan–Meier method. Differences between survival curves were tested with the log-rank test. A P value < 0.05 was considered to be statistically significant. Survival was calculated from the date of diagnosis of SCLC. BM-free survival time was calculated from the date of the diagnosis of SCLC to the date of the diagnosis of BM, and patients were censored from the moment they died.
A total of 432 patients were included, 292 of whom (68%) were staged as having ED. Table 3 shows the neurologic disorders apparent at the time of diagnosis. Disorders that were diagnosed within 4 weeks after the diagnosis of SCLC was made were considered to be present at the time of the initial diagnosis of SCLC. Seventy-four patients (18%) were found to have BM at the time of diagnosis (synchronous); in 20 patients, these BM did not appear to cause clinical signs (asymptomatic). In 11 patients, BM were found at the time of autopsy, during which SCLC was diagnosed as well. Forty-five patients did not undergo a brain scan because their clinical condition was too poor and/or they died shortly after the diagnosis of SCLC was made. For another five patients, we were unable to ascertain data regarding diagnostic procedures. Other neurologic metastatic disorders were found in 17 patients. One patient had both leptomeningeal metastases and epidural metastases. Of all nonmetastatic disorders, syndrome of inappropriate antidiuretic hormone (ADH) secretion (SIADH) was found most frequently.
|Limited “disease”||Extensive “disease”||Total||(%)|
|No. of patients||140||292||432|
|No brain metastases||140||168||308||(71)|
|Symptomatic brain metastases||43||43||(10)|
|Asymptomatic brain metastases||20||20||(5)|
|Brain metastases at autopsy||11||11||(3)|
|No brain scan performed||0||45||45||(10)|
|Other metastases (CNS) (n = 18 patients)||(4)|
|Nonmetastatic disorders (n = 37 patients)||(9)|
|Lambert–Eaton myasthenic syndrome||3||4||7|
|Subacute cerebellar degeneration||1||0||1|
The denominators in Table 4 account for all patients with LD or ED who were alive 4 weeks after diagnosis, minus those patients already diagnosed with the specific neurologic disorder involved. A total of 45 patients (10%) died within 4 weeks of the diagnosis of SCLC. As described in Table 4, 101 patients developed symptomatic BM during follow-up (metachronous). In 59 patients, it remains unknown whether they developed symptomatic BM, in part because they did not undergo a second brain scan because of their poor clinical condition or early death.
|Limited diseaseab||(%)||Extensive diseaseab||(%)|
|Symptomatic brain metastases||38/136||(28)||63/127||(50)|
|Lambert–Eaton myasthenic syndrome||3/133||(2)||0/247||(0)|
|Subacute cerebellar degeneration||0/135||(0)||0/250||(0)|
Twenty-nine patients developed leptomeningeal metastases during follow-up and 8 cases were detected at the time of diagnosis. In 23 patients with leptomeningeal metastases, BM were present as well (62%).
The majority of nonmetastatic disorders preceded the diagnosis of SCLC. SIADH was diagnosed most frequently, followed by the Lambert–Eaton myasthenic syndrome.
The median survival time in the current study group (n = 421) was 8.5 months (range, 0–154 months). The median survival of those patients treated with chemotherapy (n = 127) was 9.2 months (range, 0–154 months). As expected, survival in the patients with ED was shorter than that in the patients with LD (P < 0.0005) (Fig. 1). The median survival for patients with ED (n = 284) was 7.2 months (range, 0–124 months), whereas that for patients with LD (n = 137) was 11.9 months (range, 0–154 months). As shown in Figure 2, patients with BM only (patients with ED in whom the brain was the only site of disease dissemination) had a poorer survival than patients with LD (P < 0.0005). Patients with BM only (n = 20) had a median survival of 6.1 months (range, 0–32 months).
The 2-year cumulative risk of BM among patients who were free of BM at the time of SCLC diagnosis was 49% for patients with LD and 65% for patients with ED (Fig. 3).
Of the SCLC patients with BM, 170 died within the study period. The causes of death are described in Table 5. In 59 patients, death was considered an immediate result of BM (intracranial hypertension). The cause of death remained unknown in 45 patients who died at home. Categorized as “other” are causes such as sepsis, dehydration, and myocardial ischemia.
|No. of deceased patients||170|
|Cause of death|
During the period in which PCI was excluded from the treatment protocol (1990–1997), the 1-year cumulative incidence of BM after the diagnosis of SCLC was 33%. During the period in which PCI was applied, the 1-year cumulative incidence after the diagnosis of SCLC was 23%.
In the current study, 240 SCLC patients (56%) were diagnosed with a neurologic disorder at some point during their disease. In nearly half of the cases, the neurologic disorder already was present at the time of diagnosis. A higher frequency of BM was observed in comparison with earlier studies (18% vs. 10%).8–10 This most likely is due at least in part to differences in the diagnostic procedures utilized in the initial staging. CT scan and later MRI were used routinely in the pretreatment evaluation in the current study, which led to the identification of asymptomatic BM.
During follow-up, another 101 patients developed symptomatic BM. The lifetime incidence was lower among patients with LD compared with patients with ED.
It has been reported that BM as such do not negatively influence the prognosis for patients with SCLC.11, 12 However, we found that patients with ED who had only BM had a poorer survival than patients with LD (Fig. 2). In addition, in nearly half of the patients with BM, the cause of death was found to be related directly to the BM (Table 5). We therefore argue that BM have an adverse effect on survival in patients with SCLC.
Comparing periods in which PCI was administered with those in which it was not administered, we found that the cumulative incidence of BM was higher during the period in which PCI was not administered. This supports the findings of other authors who have reported that PCI may decrease the incidence of BM.13, 14
Leptomeningeal metastases were diagnosed for the most part during follow-up. In total, we found 37 patients with leptomeningeal metastases, which was 9% of the total patient group in the current study. Previously reported percentages have been reported to vary from 6–18%.6, 7, 15, 16
Ten patients were found to have spinal metastases at the time of diagnosis. During follow-up, another 15 patients were diagnosed with epidural metastases and 4 patients were diagnosed with intramedullar metastases. In all patients with intramedullar metastases, BM already were present. Earlier studies also reported low frequencies of spinal metastases in patients with SCLC..6, 7, 17
Because of its neuroendocrine properties, SCLC is associated more frequently with paraneoplastic syndromes compared with other malignancies.18–21 In the current study, 9% of patients were found to have a nonmetastatic paraneoplastic disorder at the time of diagnosis of SCLC. SIADH was diagnosed in 30 patients, a frequency that is similar to the findings of other investigators.19 The results of the current study confirm earlier findings, which demonstrated that paraneoplastic syndromes frequently precede the diagnosis of SCLC.22
Among the conclusions reached in the current study are: 1) at the time of diagnosis of SCLC, 18% of the patients in the current study had BM; however, these BM caused no symptoms in approximately a third of the cases; 2) contrary to reports in the literature, BM appear to have a negative effect on survival in patients with SCLC; and 3) BM are the direct cause of death in nearly half of the patients with BM.
- 3Netherlands Cancer Registry. Lung cancer and mesothelioma in the Netherlands, 1989–1997. Amsterdam: Netherlands Cancer Registry, 1997.