When lung cancer is diagnosed after a long DFS period, it is difficult to decide whether it is due to a recurrence or a secondary primary lung cancer. A 5-year DFS is an indicator of the cure of SCLC . However, in some cases of SCLC, a recurrence after a DFS period of more than 10 years has been reported [2–4]. We considered the present case as a recurrence of SCLC for a few reasons. First, the immunological staining pattern of specimens obtained in 2015 was similar to that in 2001, although synaptophysin expression was not observed in 2015. Second, recurrence is more common than a second primary SCLC in SCLC patients with a long survival period. In 60 patients with SCLC who survived for more than 5 years, nine had SCLC recurrence and none had a second primary SCLC . Johnson et al.  studied 62 SCLC patients with a DFS period of more than 2 years and reported that 18 had recurrence whereas two had a second primary SCLC. Third, considering the doubling time, a recurrence after a DFS period of more than 13 years is quite possible . Moreover, in our case, SCLC reappeared on the same side of the mediastinal lymph nodes. Therefore, we considered this case as a recurrence of SCLC. A small amount of SCLC cells had grown for 13 years. However, there remains the possibility of recurrence.
In the study of secondary PNS by Ducray et al. , it was reported that, in all patients with a cancer recurrence, a secondary PNS antedated the cancer recurrence. In our case, gait disturbance was presented before the diagnosis of an SCLC recurrence. If the patient has a recurrence or a secondary PNS, careful screening for cancer recurrence is needed.
The etiology of PNS recurrence might be mediated by onconeural antibody production stimulated by the recurrence of SCLC. In this case, the serum level of P/Q-type VGCC antibody was elevated with the recurrence of SCLC and LEMS. A previous study reported that, among the patients with PNS positive for onconeural antibody, the same antibody was detected when a secondary PNS appeared . Nagashima et al.  reported an SCLC patient with anti-Hu antibody who experienced a recurrence of SCLC with a secondary PNS. However, in contrast to our cases, those patients had a secondary PNS different from the first one [8, 9]. Anti-Hu antibody is related to some types of PNS, such as sensory neuropathy, cerebellar ataxia and limbic encephalitis . On the other hand, P/Q-type VGCC antibody was relatively specific to LEMS. The antibody was detected in 85–90% of patients with LEMS  and 10 of 12 SCLC patients with the P/Q-type VGCC antibody had LEMS . This might be owing to the difference between our case and the cases reported in literature. It is unclear how the type of PNS was determined, although the same antibody was detected. Further study is needed to understand the etiology of PNS.