Clinical outcome of patients with anti-Hu–associated encephalomyelitis after treatment of the tumor

Methods.

Since 1987, all patients with Hu-Ab diagnosed in the two participating laboratories have been included in a data base containing information on the neurologic symptoms, tumor diagnosis and staging, treatments received, and outcome of both the neurologic disorder and tumor. For this study, we selected 51 anti-Hu–positive patients who had a tumor diagnosed in life, received standard antineoplastic treatment, and who had a well-known evolution of the tumor and PEM. Table 1 summarizes the clinical characteristics of this series. The tumor response was assessed by the patient’s oncologist at the end of treatment. The neurologic disability was evaluated independently by a modified Rankin scale.3 In this scale, a score of 0 represents an asymptomatic patient; 1, symptoms do not interfere with lifestyle; 2, symptoms lead to some restriction of lifestyle but do not prevent totally independent existence; 3, symptoms significantly interfere with lifestyle or prevent totally independent existence; 4, symptoms clearly prevent independent existence, although the patient does not need constant attention; and 5, severe disability is present with total dependence requiring constant attention. A patient was considered neurologically improved or deteriorated if there was a change of at least 1 point in the scale measured at the onset and at the end of the tumor treatment.

To assess whether the SCLC of PEM patients has a more benign outcome than that of patients without PEM, the survival of patients with PEM and SCLC who did not die from treatment complications or as consequence of PEM was compared with that of a group of SCLC patients without PEM or Hu-Ab. We matched 27 pairs of patients for the following characteristics: age, performance status, tumor stage, and type of antineoplastic treatment. Because performance status in PEM patients was mostly dependent on the neurologic dysfunction, patients with PEM were divided according the presence of a Rankin score 0 to 3 or 4 to 5. Patients with Rankin score of 0 to 3 were matched with SCLC patients without PEM with Eastern Cooperative Oncology Group performance status of 0 to 2, whereas those with a Rankin score of 4 to 5 were matched with an Eastern Cooperative Oncology Group performance status of 3 to 4. In the Eastern Cooperative Oncology Group scale, 0 represents an asymptomatic patient; 1, symptomatic but ambulatory; 2, not fully ambulatory but out of bed more than 50% of the time; 3, ambulatory but in bed more than 50% of the time; and 4, bedridden.10 Patients also were matched as to whether they were treated with chemotherapy alone or if they also received thoracic radiotherapy. Those SCLC patients included who did not have PEM were selected from a group of 196 patients who entered in a previous study to evaluate the impact of Hu-Ab status in the SCLC outcome.9

Results.

Outcome of SCLC in patients with PEM.

The evolution of SCLC could be assessed in 32 of the 42 patients with PEM and SCLC who did not die from progression of PEM or other causes not related with SCLC progression. The SCLC in these 32 patients was discovered after a median delay of 5 months (range 0 to 29 months), and the tumor was symptomatic at the time of diagnosis in 8 (25%). Of the 32 SCLC patients with PEM, only 27 could be paired with a SCLC patient without PEM and negative Hu-Ab status (table 4). Five patients with PEM had asymptomatic SCLC but a bad performance status resulting from the neurologic dysfunction, and there were no SCLC patients without PEM who had a limited tumor and a similarly bad performance status to be paired. Each of the 27 patient pairs were matched exactly for age younger than 60 years or older, performance status, disease stage, site of metastasis, and treatment regimen with or without thoracic radiotherapy. Sex was matched in 25 pairs. There were no statistical differences between the different number or type of procedures to define the extent of the disease, although PEM patients had more brain CT scans than those without PEM (see table 4). Most of the patients—80% in the PEM group and 67% of those without PEM—received cisplatin-based chemotherapy. Median actuarial survival was similar in both groups (figure). However, 8 (29.6%) patients were alive in the PEM group with a median follow-up of 48 months (range 30 months to 120 months) and 2 (7.4%) in the SCLC group without PEM with a median follow-up of 63 months. Patients with PEM had a higher probability of survival at 30 months compared with those without PEM (OR 5.26; 95% CI 1.0004 to 27.6902; p = 0.03).

View this table:

• View inline
• View popup

Table 4.

Clinical characteristics of small-cell lung cancer (SCLC) patients with paraneoplastic encephalomyelitis (PEM) and those without PEM and no Hu-Ab

• Download figure
• Open in new tab
• Download powerpoint

Figure. Survival curves for small-cell lung cancer patients with (dashed line) or without (solid line) paraneoplastic encephalomyelitis. Tick marks represent patients who are alive.

Discussion.

The current retrospective study shows that in patients with PEM and Hu-Ab, complete response of the tumor to therapy is the best predictor to arrest the evolution of the neurologic dysfunction. Immunotherapy does not modify the clinical course of PEM,3,4,7 but its use was not detrimental to the tumor evolution. Last, the SCLC of PEM patients appears to have a slightly better evolution than that of patients without PEM or Hu-Ab and identical prognostic factors for survival.

Although improvement of the neurologic paraneoplastic syndrome after treatment of the tumor has been observed in isolated case reports, the general consensus is that the neurologic paraneoplastic syndrome usually runs an independent course from the tumor.14 This common belief, however, has never been assessed in a large series of patients with a known tumor response to the treatment. In 11 patients with LEMS whose SCLC was successfully treated, 7 had a sustained improvement of the neurologic syndrome.6 The results in our PEM patients agree with the findings in LEMS and emphasize the importance of achieving a complete remission of the tumor.

The 70% frequency of improvement or stabilization of PEM in the current series is higher than that described in other studies1,3,4 and probably reflects the selection of patients. The 51 patients included in this study represent less than the 30% of the patients diagnosed in the same period, and they probably had a less aggressive PEM, which allowed the patient to reach with better functional status to support the treatment of the tumor, once it was diagnosed. PEM may spontaneously arrest or improve in a few patients.15 However, this possibility is an unlikely explanation of the current results because we still observed an association between PEM stabilization and complete response of the tumor to the therapy in the subset of patients whose neurologic symptoms were clearly progressing in the month preceding the onset of the antineoplastic therapy.

Treatment of the underlying tumor cannot represent the only approach to the management of the neurologic paraneoplastic syndrome despite its beneficial effect. Many patients deteriorate in the months preceding the discovery of the tumor; PEM sometimes develops during treatment of the tumor,1 and, in at least 30% according to this series, PEM progressed even when a complete response of the tumor was achieved. These data, coupled with the possible autoimmune cause of PEM,16,17 are the rationale for the use of immunosuppressor or immunomodulating therapies. Although immunotherapy has not proved effective in most PEM patients,3,4 a few patients clearly improve,4,18 and new immunotherapy strategies probably will be tested in more PEM patients.19 A concern of this approach is that the immunosuppression could adversely affect the evolution of the tumor. Our retrospective analysis does not support this possibility, and the outcome of the SCLC was similar between PEM patients who received immunotherapy and those who did not. These data agree with those observed in patients with LEMS and SCLC6 or in children with opsoclonus and neuroblastoma.20 Unlike PEM, both neurologic syndromes improve with immunosuppressors, and the underlying tumor does not do worse. On the contrary, the occurrence of opsoclonus represents a favorable prognosis for survival in children with neuroblastoma.20

Anecdotal evidence indicates that the immune response of anti-Hu–associated PEM may control the tumor growth of the SCLC.8 The current findings show that once the SCLC is discovered and, therefore, is not completely under control of the immune system, patients with PEM do only marginally better in terms of tumor progression and survival than SCLC patients with similar prognostic factors but without PEM or Hu-Ab.21-23 The current observation agrees with our previous study, which showed that the presence of Hu-Ab in SCLC without PEM is not an independent predictor for survival.9 The current data cannot be extrapolated to the whole population of patients with PEM. We cannot exclude the fact that patients who die from the neurologic syndrome before the tumor is discovered or who are too disabled to support the antineoplastic treatment indeed have a more effective immunologic control of the underlying SCLC.

Last, although the SCLC was diagnosed in an asymptomatic stage in most PEM patients, the chances of curing the tumor still are suboptimal, and only 30% survive more than 2 years. This poor result probably is explained by the finding that the SCLC of PEM patients usually presents mediastinal adenopathies at the time of the diagnosis,24 the staging situation most commonly found in SCLC patients without PEM whose tumor is limited to the chest.25 This observation, coupled with the poor effect of immunotherapy in PEM and the association between tumor response and PEM stabilization, further supports the aggressive search for the tumor in these patients. When the tumor is not found at the initial evaluation of PEM patients who have a high risk for SCLC, who are older than 50 years, and have a smoking habit, the possibility of treatment with chemotherapy to stop the progression of the neurologic disorder and to cure the still-microscopic SCLC should be strongly considered in the setting of a prospective therapeutic protocol.