bs_bs_banner

ORIGINAL ARTICLE

Paraneoplastic Hu and CRMP5 antibodies are present in smokers without cancer or neurological disease TOR H. QVALE,1 ANETTE STORSTEIN,2 KIBRET MAZENGIA,1 TOMAS M.L. EAGAN,3,4 PER S. BAKKE3,4 AND CHRISTIAN A. VEDELER1,2 1

Department of Clinical Medicine and 3Clinical Science, University of Bergen, Departments of 2Neurology and 4Thoracic Medicine, Haukeland University Hospital, Bergen, Norway

ABSTRACT Background and objective: We investigated if the paraneoplastic Hu and collapsin response mediator protein 5 (CRMP5) antibodies could be used as early markers for lung cancer in smokers with or without chronic obstructive pulmonary disease (COPD). Methods: Hu and CRMP5 antibodies were measured by radioimmunoprecipitation assay (RIPA) in sera from 552 smokers; 379 with and 173 without COPD. Three hundred blood donors served as controls. The positive sera were also tested by indirect immunofluorescence and line blot with recombinant proteins. The 552 smokers were matched with data from the Cancer Registry of Norway, and the hospital medical records from the subjects positive for Hu and CRMP5 antibodies were reviewed. The mean follow-up time was 4.4 years (range 2.5–5.7 years). Results: The RIPA showed that 5/379 (1.3%) smokers with COPD had Hu antibodies and 1/379 (0.3%) smokers with COPD had CRMP5 antibodies. Only the smoker with the highest RIPA index had Hu antibodies also detected by immunofluorescence and line blot. One of 173 (0.6%) smokers without COPD had Hu antibodies, but none had CRMP5 antibodies. None of the 300 controls had Hu antibodies, but 2/300 (0.7%) had CRMP5 antibodies. Hu antibodies remained positive for more than 5 years. No cancer or neurological disease was recorded in the Hu or CRMP5 positive patients. The total cancer frequency in the smokers with and without COPD was 70/552 (13%). Conclusions: Hu and CRMP5 antibodies were not associated with cancer or neurological disease in a large cohort of smokers and are therefore not always paraneoplastic. Key words: chronic obstructive, nervous system, paraneoplastic syndromes, pulmonary disease, smoking. Abbreviations: BCCS, Bergen COPD Cohort Study; BSA, bovine serum albumin; COPD, chronic obstructive pulmonary Correspondence: Christian A. Vedeler, Department of Neurology, Haukeland University Hospital, Jonas Lies vei 65, 5021 Bergen, Norway. Email: [email protected] Received 3 December 2013; invited to revise 15 December 2013; revised 3 February 2014; accepted 6 February 2014 (Associate Editor: Chi Chiu Leung). Article first published online: 3 April 2014 © 2014 Asian Pacific Society of Respirology

SUMMARY AT A GLANCE Paraneoplastic Hu and CRMP5 antibodies were tested in a large cohort of smokers. Such antibodies were found in smokers without cancer or neurological disease and are therefore not always paraneoplastic.

disease; CRMP5, collapsin response mediator protein 5; CT, computed tomography; FEV1, forced expiratory volume in 1 s; GOLD, Global Initiative for Chronic Obstructive Lung Disease; IF, immunofluorescence microscopy; ITT, in vitro transcription and translation; NSCLC, nonsmall cell lung cancer; PAS, protein A-sepharose; RIPA, radioimmunoprecipitation assay; RT, room temperature; SCLC, small cell lung cancer.

INTRODUCTION Paraneoplastic neurological syndromes are immunemediated, remote effects of cancer, which manifest in less than 1% of cancer patients, often before a tumour is found.1 These syndromes arise from an immunological cross-reaction between antigens common to the cancer cells and cells in the nervous system.1 Lung cancer is associated with smoking and chronic obstructive pulmonary disease (COPD). Lung cancer is also the tumour type most commonly associated with paraneoplastic neurological syndromes.2 Several paraneoplastic antibodies associated with paraneoplastic neurological syndromes have been characterized.3 Hu and collapsin response mediator protein 5 (CRMP5) antibodies are among the most common paraneoplastic antibodies and are usually associated with lung cancer and paraneoplastic encephalomyelitis, limbic encephalitis or subacute sensory neuronopathy.2,3 We have found that a radioimmunoprecipitation assay (RIPA) is more sensitive than immunohistochemistry and immunoblot in detecting paraneoplastic antibodies.2,4 By using RIPA, Hu and CRMP5 antibodies have been found in Respirology (2014) 19, 730–734 doi: 10.1111/resp.12292

Paraneoplastic antibodies in smokers

22.5% and 5%, respectively, in patients with small cell lung cancer (SCLC) and in 0% and 0.7%, respectively, in blood donors.4 Paraneoplastic antibodies have so far not been studied in smokers or patients with COPD. Such antibodies are used as markers for paraneoplastic neurological syndromes, but the antibodies can also be present in cancer patients without neurological disease.4 The aim of this study was to investigate if the paraneoplastic Hu and CRMP5 antibodies could be used as early markers for lung cancer in smokers with or without COPD. Hu and CRMP5 antibodies were chosen because they are among the most common paraneoplastic antibodies detected in lung cancer.

METHODS Subjects The patient cohort consisted of 552 smokers, 379 with COPD and 173 without COPD, from the Bergen COPD Cohort Study (BCCS), an ongoing cohort study from Western Norway. Age at enrolment was 40–75 years, and all had an at least 10 pack-years smoking history. Smokers with inflammatory disorders and active cancer in the last 5 years were excluded. The study design, selection of participants and sampling of clinical data have been described previously.5 Mean follow-up time of the 552 smokers was 4.4 years (range 2.5–5.7 years). The serum samples were collected every 6 months. Chest computed tomography (CT) scans were performed at baseline, 12 and 36 months. The data were linked to the Cancer Registry of Norway. Hospital medical records and chest CT scans were reviewed for symptoms and signs that could suggest neurological disease or lung cancer for the Hu or CRMP5 antibody-positive patients. Sera from 300 adult blood donors served as healthy controls for the Hu and CRMP5 antibody tests. The blood donors did not have a clinical diagnosis of COPD or other diseases. However, their smoking habits and lung function were unknown. The study was approved by the regional ethics committee (REKVEST) of Western Norway. Antibody detection Sera from all 552 smokers drawn at the initial study visit were tested by RIPA as described previously.6 Bluescript plasmids with Hu and CRMP5 were separately transfected into Escherichia coli XL-1 via electroporation. Plasmids were purified with a Qiagen plasmid maxi kit (Qiagen, Hilden, Germany). Transcription and translation (ITT) were performed in vitro using a TNT T7 coupled reticulocyte lysate system (Promega, Madison, WI, USA) for the generation of 35S-labelled Hu and a TNT T3 coupled system for the generation of 35S-labelled CRMP5. Microtiter plates (MABV N12, Merck Millipore, Bedford, MA, USA) were preincubated with 200 μL buffer A (150 mM NaCl, 20 mM Tris-HCl, 0.01% azide with pH = 8.0) per well for 1 h at room temperature (RT). Buffer A was discarded and 200 μL buffer A containing 1% bovine serum albumin (BSA) (w/v) was added per © 2014 Asian Pacific Society of Respirology

731 well and incubated for 2 h. The wash was discarded and the wells were washed twice with 200 μL buffer A containing 0.05% Tween-20 and once with buffer B (0.1% w/v BSA, 0.05% mL Tween-20 in Buffer A). Sera and ITT product were added in triplicates to 96-well mixing plates by adding 5 μL patient serum to 45 μL buffer B with 1.04 μL ITT product and incubated overnight at 4°C. A pool of blood donors were used as negative control and sera from Hu/CRMP5immunized rabbits were used as positive control. Protein A-sepharose (PAS) was prepared by adding 4 mL PAS to 9 mL ice cold buffer B in a 15-mL Falcon tube, mixing, centrifuging at 188 g in 5 min, resuspending the PAS in 15 mL buffer B, centrifuging again and resuspending PAS to a total volume of 6 mL. The immunoprecipitation was performed by placing microtiter plates prepared as described above on ice. The plates were washed in a vacuum manifold (1225 Sampling Manifold, Merck Millipore) by first adding 200 μL buffer B per well three times with suction, then adding 150 μL buffer B without suction and incubating with shaking at 120 rpm for 5 min. This procedure was repeated once before the plates were washed two last times with 200 μL buffer B. The filters were dried at RT overnight, put in a plastic container and 20 μL scintillation fluid (Microcint-0, PerkinElmer, Boston, MA, USA) were added and the plates were covered with plastic film. The plates were counted in a β-counter (Topcount NXT, PerkinElmer). The indexes for the samples were calculated from the counts per minute value by the same method as described previously.6 The mean index of the controls added to three standard deviations of it was the cut-off value used to discriminate between positive and negative samples. The cut-off used for the Hu and CRMP5 index was 262 and 216, respectively. Serum samples from participants with positive RIPA index and the seven patients with lung cancer were also tested by line blot and indirect immunofluorescence microscopy (IF). The Ravo PNS+2 Blot (detects HuD, Yo, Ri, CRMP5, Amphiphysin, Ma1, Ma2, SOX1 and GAD65 antibodies) from Ravo Diagnostika (Freiburg, Germany) was used according to the manufacturer’s instructions. The IF test was performed by applying sera on fixed rat cerebellum/ brain stem slices. Slices were first blocked with 25 μL blocking buffer (phosphate buffered saline, Triton X-100 and BSA) and then incubated with patient serum at dilution 1:100 and 1:500 in blocking buffer at 4°C overnight. The slices were then washed in blocking buffer and incubated with 30 μL Alexa anti-human 488 (Invitrogen A11013, Leiden, The Netherlands) diluted 1:400 in blocking buffer at RT for 90 min. After washing, the samples were visualized with a fluorescence microscope (model DM IL; Leica Microsystems GmbH, Wetzlar, Germany). Samples with staining of the nucleus without staining of the nucleolus were interpreted as Hu positive.

Statistical analysis Comparison of pack-year and age means for smokers/COPD relative to controls were performed using the Welch two-sample t-test (function: t.test), Respirology (2014) 19, 730–734

732

TH Qvale et al.

which is included in the statistical application R v. 2.15 (R. Foundation, Vienna, Austria) on a MacBook running OS X 10.8 (Apple Inc., Cupertino, CA, USA). Comparison of RIPA means for smokers/COPD relative to controls was also performed with the same test. A two-sided Fisher’s exact test was done using the built-in fisher.test function in R to compare the number of antibody positive in the COPD and nonCOPD smoker and blood donor groups. Nonparametric correlations and one-way analysis of variance (significance level 0.05) among Hu/CRMP5 indexes, lung function, the number of pack-years smoked and Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage of participants with COPD were done using SPSS 21 (IBM Corp., Armonk, NY, USA).

RESULTS Clinical characteristics The characteristics of smokers, with and without COPD, blood donors and the results of the antibody assays are shown in Table 1. Estimated pack-years of smoking and forced expiratory volume in 1 s (FEV1) per cent predicted are from the baseline visit of the BCCS. Both age and the number of pack-years smoked differed significantly (P < 0.01; t-test) between smokers with and without COPD. The proportion of men was lowest among smokers without COPD and did not differ between the other two groups. The smokers without COPD had an average FEV1 in per cent predicted comparable with a reference population. The COPD patients were severely impaired at inclusion, as 178 (47%) were GOLD stage 2, 162 (43%) were GOLD stage 3 and 39 (10%) were GOLD stage 4. Seventy of the 552 smokers had a history of cancer of which 27 were given the diagnosis during the study period. Seven subjects, all of them were given the diagnosis after the study start, had lung cancer, six

Table 1 Characteristics of demography, smoking exposure and paraneoplastic antibodies in smokers with and without chronic obstructive pulmonary disease (COPD) and in blood donors Smokers Smokers Blood with COPD without COPD donors n 379 Males, % 60 Mean age 64 Mean pack-years 41 Hu antibodies, n 5 CRMP5 antibodies, n 1 48.9 (14.1) Mean (SD), FEV1 in % predicted Lung cancer, n 7 Nonlung cancer, n 50

173 53 53 30 1 0 102.6 (9.14) 0 13

300 62 45 NA 0 2 NA 0 0

FEV1, forced expiratory volume in 1 s; NA, not applicable; SD, standard deviation. Respirology (2014) 19, 730–734

with non-SCLC (NSCLC) and one with neuroendocrine carcinoma with no specific histology. All of the lung cancer patients had COPD.

Antibodies The RIPA detected a positive Hu index in five of the 379 smokers with COPD and one of the 173 smokers without COPD (Fig. 1a). One of the 379 smokers with COPD and two of the 300 blood donor controls had a positive CRMP5 index (Fig. 1b). Hu antibodies were not significantly more frequent in smokers with COPD than smokers without COPD or blood donors (P = 0.67 and P = 0.07; respectively; Fisher’s exact test). However, the mean Hu or CRMP5 antibody index for the smokers with or without COPD was slightly lower compared with the blood donors (P < 0.01; t-test) (Fig. 1). Only the patient sample with the highest Hu index (initial index = 1400) gave a positive result with indirect immunofluorescence and line blot assays. None of the Hu- or CRMP5-positive patients or lung cancer patients had other autoantibodies measured by line blot. Two of five smokers with COPD with positive Hu index on the initial sample remained positive on all samples (Fig. 2). No additional samples were available for the smoker without COPD with an initial positive Hu sample. In none of the Hu- or CRMP5positive smokers was cancer reported in the cancer registry or in their medical records, which included regular chest CT scans on all subjects. There was no significant correlation between the Hu and CRMP5 indexes and lung function, the number of pack-years smoked or GOLD stage in smokers with COPD.

DISCUSSION This study showed that Hu or CRMP5 antibodies can occur in smokers with COPD without the presence of cancer or paraneoplastic neurological syndrome. The smoker with the highest Hu index was also positive using line blot and immunofluorescence tests, and we have shown previously that the RIPA assay is more sensitive than other assays.2 In the two smokers with the highest Hu antibody levels, the Hu index remained positive for more than 5 years. Hu and CRMP5 antibodies are usually associated with paraneoplastic neurological syndromes such as encephalomyelitis, limbic encephalitis or subacute sensory neuronopathy2; we have previously shown that these antibodies also can be detected in cancer patients without neurological disease.4 In this study, we demonstrate that Hu and CRMP5 antibodies are not always associated with paraneoplastic neurological syndromes. Furthermore, we found that Hu and CRMP5 antibodies do not correlate with lung function, pack-years or GOLD stage. Recently, seven autoantibodies (p53, NY-ESO-1, CAGE, GBU4–5, SOX2, HuD and MAGE A4) have been shown to detect early cases of lung cancer.7 This panel of antigens detects antibodies using an enzymelinked immunosorbent assay-based principle. The kit © 2014 Asian Pacific Society of Respirology

Paraneoplastic antibodies in smokers

733

Figure 1 Boxplot of radioimmunoprecipitation assay (RIPA) indexes. The plots show the median, quartiles and whiskers of the largest/smallest value that is 1.5× the distance from the quartiles. Horizontal lines separate positive and negative indexes. (a) Hu indexes of smokers, chronic obstructive pulmonary disease (COPD) patients and controls. (b) CRMP5-indexes of smokers, COPD patients and controls.

Figure 2 Follow-up samples from patients with a positive radioimmunoprecipitation assay (RIPA) Hu index at first visit. The first visits were in 2006/2007, and the last visits were in 2011/ 2012. The horizontal black line denoted ‘cut-off’ separates ) 3329, ( ) 3475, positive and negative samples. Patient ID: ( ) 3644, ( ) 3330, ( ) 3649. (

is made for the early detection of both NSCLC and SCLC. The sensitivity for lung cancer was highest when all seven antibodies were tested in combination. In this study, we did not find that screening for Hu or CRMP5 antibodies could predict lung cancer in smokers or COPD patients. Our findings imply that lung cancer screening using various kits could generate false-positive results. The mean Hu or CRMP5 indexes for the smokers were slightly lower than for the blood donors. The reason for this could be that chronic smoking suppresses adaptive immunity, which reduces immunoglobulin levels and altered levels of autoantibodies.8 © 2014 Asian Pacific Society of Respirology

The prevalence of paraneoplastic antibodies are not related to age,2 which differed among the study groups. It has been shown that smoking is associated with double-stranded DNA antibody production in systemic lupus erythematosus9 and citrullinated protein antibodies in rheumatoid arthritis.10 The results of our study show that smoking can also trigger the formation of other autoantibodies, namely anti-Hu and antiCRMP5. It is therefore possible that smoking may expose antigens in genetically disposed persons that induce an immune response. Smoking has been shown to be a risk factor for autoimmune diseases, such as rheumatoid arthritis10 and multiple sclerosis.11 Seventy of the 552 smokers had a history of cancer, but only seven of these involved the lung and none of these patients had Hu or CRMP5 antibodies. This can partly be explained by the fact that none of the seven patients with lung cancer had SCLC, the cancer form most often associated with Hu and CRMP5 antibodies. We have previously found Hu and CRMP5 antibodies in 22.5% and 5%, respectively, in patients with SCLC.4 In a SCLC mouse model, it was found that 14% of the mice had Hu antibodies and that the antibody response predated the development of lung cancer by about 100 days,12 suggesting that the smokers with Hu antibodies could be in the process of carcinogenesis. This was not confirmed in our five patients with Hu antibodies followed for more than 5 years with chest CT scans. Fluorodeoxyglucosepositron emission tomography was not performed in these patients as this currently has its place in the follow up of lung nodules first detected on CT.13 In conclusion, our study shows that testing for Hu and CRMP5 antibodies alone does not predict lung cancer. Screening for paraneoplastic antibodies should only be performed in patients with suspected paraneoplastic neurological syndrome. Respirology (2014) 19, 730–734

734

TH Qvale et al.

Acknowledgements The authors thank Mette Haugen for technical assistance. The study was supported by grants from the University of Bergen, Norway.

7

REFERENCES

8

1 Darnell RB, Posner JB. Paraneoplastic syndromes involving the nervous system. N. Engl. J. Med. 2003; 349: 1543–54. 2 Storstein A, Monstad SE, Haugen M, Mazengia K, Veltman D, Lohndal E, Aarseth J, Vedeler C. Onconeural antibodies: improved detection and clinical correlations. J. Neuroimmunol. 2011; 232: 166–70. 3 Graus F, Delattre JY, Antoine JC, Dalmau J, Giometto B, Grisold W, Honnorat J, Smitt PS, Vedeler C, Verschuuren JJ et al. Recommended diagnostic criteria for paraneoplastic neurological syndromes. J. Neurol. Neurosurg. Psychiatry 2004; 75: 1135–40. 4 Monstad SE, Knudsen A, Salvesen HB, Aarseth JH, Vedeler CA. Onconeural antibodies in sera from patients with various types of tumours. Cancer Immunol. Immunother. 2009; 58: 1795–800. 5 Eagan TM, Ueland T, Wagner PD, Hardie JA, Mollnes TE, Damas JK, Aukrust P, Bakke PS. Systemic inflammatory markers in COPD: results from the Bergen COPD Cohort Study. Eur. Respir. J. 2010; 35: 540–8. 6 Monstad SE, Drivsholm L, Storstein A, Aarseth JH, Haugen M, Lang B, Vincent A, Vedeler CA. Hu and voltage-gated calcium

Respirology (2014) 19, 730–734

9

10

11

12

13

channel (VGCC) antibodies related to the prognosis of small-cell lung cancer. J. Clin. Oncol. 2004; 22: 795–800. Chapman CJ, Healey GF, Murray A, Boyle P, Robertson C, Peek LJ, Allen J, Thorpe AJ, Hamilton-Fairley G, Parsy-Kowalska CB et al. EarlyCDT(R)-Lung test: improved clinical utility through additional autoantibody assays. Tumour Biol. 2012; 33: 1319–26. Sopori M. Effects of cigarette smoke on the immune system. Nat. Rev. Immunol. 2002; 2: 372–7. Freemer MM, King TE Jr, Criswell LA. Association of smoking with dsDNA autoantibody production in systemic lupus erythematosus. Ann. Rheum. Dis. 2006; 65: 581–4. Haj Hensvold A, Magnusson PK, Joshua V, Hansson M, Israelsson L, Ferreira R, Jakobsson PJ, Holmdahl R, Hammarstrom L, Malmstrom V et al. Environmental and genetic factors in the development of anticitrullinated protein antibodies (ACPAs) and ACPA-positive rheumatoid arthritis: an epidemiological investigation in twins. Ann. Rheum. Dis. 2013. Hedstrom AK, Hillert J, Olsson T, Alfredsson L. Smoking and multiple sclerosis susceptibility. Eur. J. Epidemiol. 2013; 28: 867– 74. Kazarian M, Calbo J, Proost N, Carpenter CL, Berns A, Laird-Offringa IA. Immune response in lung cancer mouse model mimics human anti-Hu reactivity. J. Neuroimmunol. 2009; 217: 38–45. Henschke CI, Yankelevitz DF, Libby DM, Pasmantier MW, Smith JP, Miettinen OS. Survival of patients with stage I lung cancer detected on CT screening. N. Engl. J. Med. 2006; 355: 1763–71.

© 2014 Asian Pacific Society of Respirology

Paraneoplastic Hu and CRMP5 antibodies are present in smokers without cancer or neurological disease.

We investigated if the paraneoplastic Hu and collapsin response mediator protein 5 (CRMP5) antibodies could be used as early markers for lung cancer i...
338KB Sizes 0 Downloads 0 Views