This is the first examination of anti-Hu reactivity in relation to lung cancer risk in a population-based study. To sensitively determine levels of anti-Hu reactivity in SCLC cases and control subjects, we used recombinant HuD protein in a Western blot analysis, using Biorad Quantity One 1-D Analysis Software to quantitate 5-min exposures in a blinded fashing. Only reactivity against the correct molecular weight band was scored as positive. This assay allows the sensitive and specific unbiased detection of even low titers of anti-Hu antibodies while minimizing false positives that may be obtained with other methods [35
]. For example, as was done in the Dalmau study, the use of crude neuronal extracts may lead to scoring of sera with reactivity against proteins other than Hu [21
], while in ELISA the size of the reacting protein is not visible. The use of imaging software likely increased our sensitivity compared to other studies, which often score samples using a more subjective scals (e.g. “mild, moderate, intense”) (see below).
In our study, we did not establish a ‘reactivity’ cut point based on reactivity levels in control subjects. Instead our goal was to measure the direct level of anti-Hu reactivity in all subjects and evaluate associations between level of anti-Hu reactivity and SCLC. In contrast. Dalmau and colleagues considered reactivity of normal human sera (equivalent to 6U/ml) as background; sera from SCLC patients with and without PEM/SN were considered positive only if their reactivity was greater than twice the background value found in normal sera. Seven of 44 (16%) SCLC patients without PEM/SN had detectable levels of anti-Hu antibody (average 76 U/ml) [21
]. Other studies that used recombinant HuD antigen on Western blots created categories of mild, moderate, or intense to describe anti-Hu reactivity [10
]. These studies identified positive reactivity in 16% of their total SCLC population [10
], and 17% of SCLC cases [22
]. Guidelines for detection of anti-Hu antibodies established by the International Society of Neuro-Immunology noted that specimens yielding no reactivity on immunohistochemistry may nonetheless be detected by Western blot analysis using highly affinity purified recombinant proteins, suggesting that these methods may be more sensitive [35
]. When using a highly sensitive in vitro transcription-translation (ITT) assay and immunoprecipitation, Monstad et al, observed anti-Hu reactivity in 25.5% of SCLC patients. When they submitted the same specimens to a dot blot analysis using recombinant HuD and 1:2000 dilution, they found results (18.5%) comparable to previous studies [23
Methods for classifying reactivity and detection methods thus appear to impact prevalence of anti-Hu reactivity. To make compare our analysis to other studies, we could consider the highest anti-Hu reactivity value in the non-smokers as our baseline cutoff value (~7500). In cutting our reactivity at 7500 and considering values above 7500 as positive, we found that 15% of SCLC patients had detectable antibody levels. This finding is similar to previous results [10
], but, in our opinion, negates the important finding that anti-Hu reactivity is detectable at values lower than 7500 among smoking and non-smoking controls.
We found a moderate association between anti-Hu reactivity and SCLC. The odds ratio for moderate to high positive anti-Hu reactivity (>1800 units), adjusted for total pack years of smoking, was marginally statistically significant (odds ratio = 3.2 (95% confidence interval = 0.98 to 13.4)).
Examining the frequency of detectable anti-Hu reactivity, we observed an unadjusted association between anti-Hu reactivity and case-control status (p = 0.01). We adjusted for total pack years of smoking using analysis of covariance (ANCOVA), and the association between continuous anti-Hu reactivity and case-control status was significant (p = 0.05). We compared anti-Hu reactivity among the controls according to smoking status. We found no overall difference and no differences when we stratified by gender and race. In the retrospective studies by Dalmau and Verschuuren, there was a predominance of women who were anti-Hu positive [21
]. Graus et al observed no difference in mean age and sex in SCLC patients with and without PEM/SN [10
]. When comparing patients positive or negative for anti-Hu antibodies, Monstad et al reported no significant differences between sex or age [23
]. These studies did not however examine anti-Hu reactivity according to race.
We observed that anti-Hu reactivity was protective for survival from small cell lung cancer in our study, however results can only be considered suggestive due to small numbers. Our results are in agreement with the previously observed longer survival time of anti-Hu positive SCLC patients [22
], the more limited disease observed in anti-Hu positive SCLC cases [21
], and reports about small cell lung cancer cases with paraneoplastic sensory neuronopathy in complete remission [40
]. The presence of anti-Hu antibodies may indicate an immune system response to small cell lung cancer that could function to prolong survival.
Our study differs from previously conducted research on anti-Hu antibodies and SCLC in that we used healthy population-based controls matched on age, race, sex, and smoking status. Large scale studies are needed to determine whether anti-Hu reactivity profiles can serve as an early detection marker for SCLC. Interestingly,, one of the smoking controls with a detectable level of anti-Hu antibody, healthy at time of study participation, died of SCLC several years after the study was concluded. This observation supports a possibility that moderate level anti-Hu reactivity might be an indicator for future SCLC. While this is an anecdotal occurrence, appearance of a small cell lung cancer case in a control group of this size would not be expected given the underlying background of SCLC incidence in Los Angeles County that occurred during the study period [3
We found that smoking was not associated with anti-Hu reactivity among our healthy controls. Smoking greatly increases the risk of small cell lung cancer. Development of anti-Hu reactivity, on the other hand, may signal the presence of a small cell lung tumor, but it is not necessarily related to smoking.
Small cell lung cancer is one of the most rapidly fatal cancers [3
]. Identification of high-risk smokers that may develop SCLC could help to improve survival. Smokers who are determined to be at risk for SCLC could be closely monitored and could serve as candidates for chemoprevention once effective agents become available. In addition, they should strongly be encouraged to quit smoking. Although only suggestive, given our small sample size, reactivity against Hu proteins and other auto-antigens could indicate the first signs of a growing small cell lung tumor, potentially in time to prolong survival.