Article Text
PDF
Statistics from Altmetric.com
Introduction
The appearance of a paraneoplastic neurological disorder (PND) with Hu-antibodies, often before tumour detection, together with spontaneous tumour regression in a few patients, suggests that immune responses to HuD, possibly mediated by cytotoxic T-cells, could be beneficial to the patient.1 Nevertheless, in previous retrospective studies, survival in Hu-antibody-associated PNDs (Hu-Ab/PNDs) in general was similar to those in patients with small-cell lung cancer (SCLC) without neurological illness.2 3
Here we studied, for the first time prospectively, tumour outcomes in a large series of patients with Hu-Ab/PNDs, comparing with contemporaneous patients with SCLC without neurological symptoms.
Methods
Patients
Written consent was obtained from all patients. From July 2012 to June 2015 prospectively we collected all serum samples that were positive for Hu-antibodies at Oxford University Hospitals as part of routine diagnostic testing via referral from any of England’s 152 acute hospital trusts. Diagnostic categorisation of associated PNDs was agreed between the research team and the referring physician, in line with published diagnostic criteria.4 For survival comparisons, contemporaneous control data and pre-treatment serum samples were obtained prospectively from 245 patients with SCLC, without PNDs, from the Nottingham Trent region over the same time period. Serum samples were stored at −80°C and retested for Hu and SOX2 antibodies by ELISA, and paraneoplastic antibodies tested on frozen rat cerebellum with positive results confirmed using a multi-line blot (Ravo Diagnostika, Freiburg, Germany) (see online supplemental methods).
Supplemental material
Statistical analysis
Survival statistics from date of SCLC diagnosis in PND and control groups were calculated using log-rank tests, and differences in cancer survival according to presence of PNDs and Hu-antibodies were evaluated using a Cox proportional hazards model.
Results
Patients with Hu-antibody PNDs and SCLC controls
One hundred and three patients were recruited. Age at diagnosis was similar between the groups but there were more women among the 103 Hu-Ab/PND patients than in the 245 SCLC controls (60% vs 49%; p=0.048). The most frequent pure clinical syndromes were subacute sensory neuronopathy, cerebellar degeneration/ataxia or limbic encephalitis, with multifocal involvement, autonomic failure, Lambert-Eaton myasthenic syndrome or neuropathy (sensorimotor, pure motor) in the remaining patients (online supplemental table 1). Fourteen of the 103 patients with Hu-Ab/PNDs had no cancer identified and 15 had tumours other than SCLC, leaving 64 patients with Hu-Ab/PND/SCLC (online supplemental table 1). Limited stage SCLC at cancer diagnosis was more common in patients with Hu-Ab/PND/SCLC (32/64, 50%) compared with SCLC controls (88/245, 35.9%) (p=0.04). Chemoradiotherapy regimens were similar in the patients with Hu-Ab/PND/SCLC compared with the SCLC control patients (online supplemental table 2).
Hu-antibody titres were higher in the patients with Hu-Ab/PND/SCLC than in the 22/245 SCLC control patients positive for Hu-antibodies (p=0.018), and higher in study patients with cancer (n=89) compared with those without detectable cancer on follow-up (n=14) (p=0.019). SOX2-antibodies were more common in the patients with Hu-Ab/PND/SCLC (31/60 tested, 52%) than in those with other tumours or none (6/37 tested (16%); p=0.0005) and predicted SCLC (positive predictive value 85%, negative predictive value 53%).
Survival of patients with PND and SCLC controls
At the end of study follow-up, fewer patients with Hu-Ab/PNDs in the non-cancer group (8/14, 57%) had died compared with patients with Hu-Ab/PNDs with SCLC (49/64, 77%) and those with other cancers (19/25, 76%). Median survival from neurological symptom onset was favourable in patients with Hu-Ab/PND without cancer (28 months) compared with the patients with cancer (13.5 months; p=0.02; figure 1A).
Kaplan-Meier survival curves. (A) Patients with Hu-Ab/PND without cancer showed improved median survival from neurological symptom onset (28 months) compared with patients with Hu-Ab/PND with cancer (13.5 months; p=0.02, Log rank). (B) Patients with small-cell lung cancer (SCLC), Hu antibodies and paraneoplastic syndrome (Hu-Ab PND and SCLC) show improved tumour survival compared with patients with SCLC alone (p=0.04, Log rank). Hu-Ab/PND, Hu-antibody-associated PND.
Cause of death was considered to be cancer-related in the majority (45/76 deaths overall; 34/49 deaths in patients with Hu-Ab/PND/SCLC; 11/19 deaths in patients with Hu-Ab/PNDs with other cancers), particularly in patients with Hu-Ab/PND/SCLC when the PND was confined to the peripheral nervous system (20/29, 69%). Among eight deaths in the 14 Hu-Ab/PND non-cancer patients, five were due to a PND affecting the central nervous system and three with peripheral nervous system involvement died of disease unrelated to the PND (cerebral haemorrhage, pneumonia, cardiac failure). Six of the eight had died without postmortem (after >18 months follow-up tumour surveillance in five), and in two there was no evidence of cancer even at postmortem. However, the remaining six were still alive at last review after a median of 54 months (range 28–126 months).
Median cancer survival of all 64 patients with Hu-Ab/PND/SCLC from time of SCLC diagnosis (11.5 months) was greater than in the 245 SCLC control patients (9.75 months; p=0.04; figure 1B), but not significantly different from the 25 patients with Hu-Ab/PNDs with tumours other than SCLC (7 months; p=0.7, Log rank; online supplemental table 1). However, when allowing for known SCLC survival prognostic factors of disease extent, age and sex in multivariate Cox regression analysis, the presence of Hu-antibody-associated PND with SCLC did not confer a significant survival advantage (p=0.24, HR 0.83, 95% CI 0.61 to 1.13; online supplemental table 3). Nevertheless, a higher proportion of patients with Hu-Ab/PND/SCLC survived longer than 48 months (10/64 (16%)) compared with patients with SCLC without neurological symptoms (16/245 (6.5%); p=0.02). Patients with Hu-Ab/PNDs who remained free of cancer at 18 months were likely to remain cancer-free (positive predictive value 95.5%, negative predictive value 92.9%).
Discussion
The clinical course, tumour associations, onconeural antibodies and survival were studied in the first large prospective series of unselected patients with Hu-antibody-associated PNDs, comparing, without bias, with those of a contemporaneous control group with SCLC. The results showed a difference in median survival time between patients with Hu-Ab/PND/SCLC and patients with SCLC, although multivariate analysis suggested that the presence of Hu antibodies and PNDs were not independent prognostic risk factors for survival in SCLC. Previous large, retrospective studies of patients with Hu-Ab/PNDs have described similar survival times (7–11.8 months, irrespective of tumour type),3 5 6 but in none of these series was the survival of patients with Hu-Ab/PND/SCLC compared with SCLC control patients with .
Our findings suggest that patients with Hu-Ab/PND/SCLC do have a modest overall favourable prognosis compared with patients with SCLC without neurological illness, with more frequent limited stage SCLC association, consistent with the important role of lead-time bias in a group where the Hu-Ab/PND presentation triggers high vigilance cancer screening.
Conclusion
We have defined the clinical course and tumour associations in a broad, unselected cohort of patients of all ages with anti-Hu antibody-associated PNDs, where prospective data collection has allowed us to demonstrate that survival in patients with Hu-Ab/PND may be due to lead-time bias, rather than independent effects of the immune system on the tumour.
Ethics statements
Patient consent for publication
Ethics approval
The study was approved and registered by our institutional review board (NRES Committee East Midlands—Leicester (10/H0406/94)).
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors PM contributed to study conception, data collection, design, analysis, writing of the first draft and revising the manuscript for important intellectual content. BL, ST, TM, PG, CJC, VB, BF and AV contributed to data collection, analysis, review of the first draft and revising the manuscript for important intellectual content.
Funding PG was supported by a fellowship grant from Myaware (UK).
Competing interests Bethan Lang and Angela Vincent are co-applicants and receive royalties on patent application WO/2010/046716 entitled ‘Neurological Autoimmune Disorders’. The patent has been licensed to Euroimmun AG for the development of assays for LGI1 and other VGKC-complex antibodies. All other authors have no conflicts of interests.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.