Data on the utilization of paraneoplastic syndrome autoantibody testing at an academic medical center

Paraneoplastic syndromes are rare conditions associated with characteristic autoantibodies produced by malignancy, although similar autoantibodies and clinical presentations may occur in the absence of any neoplasm. Testing for paraneoplastic syndromes often involves panels of autoantibody assays. While autoantibody testing may reveal or confirm actionable clinical diagnoses, inappropriate utilization of testing may be low yield and further lead to false positives that may confuse the clinical picture. There is thus opportunity to improve patient care by analyzing patterns of paraneoplastic autoantibody test utilization. The data in this article provides results from detailed retrospective review of patients tested by 7 autoantibody tests or test panels offered by two large reference laboratories in the United States. The data include 1,446 tests performed on 1,338 unique patients at an academic medical center. For all results, detailed chart review revealed main category of presenting symptoms, patient location at time of testing (either inpatient or outpatient), sex, age, whether cancer was present at the time of testing or later detected, and the specific results of the testing. The data are summarized by category of testing and specific autoantibodies.

Paraneoplastic syndromes are rare conditions associated with characteristic autoantibodies produced by malignancy, although similar autoantibodies and clinical presentations may occur in the absence of any neoplasm. Testing for paraneoplastic syndromes often involves panels of autoantibody assays. While autoantibody testing may reveal or confirm actionable clinical diagnoses, inappropriate utilization of testing may be low yield and further lead to false positives that may confuse the clinical picture. There is thus opportunity to improve patient care by analyzing patterns of paraneoplastic autoantibody test utilization. The data in this article provides results from detailed retrospective review of patients tested by 7 autoantibody tests or test panels offered by two large reference laboratories in the United States. The data include 1,446 tests performed on 1,338 unique patients at an academic medical center. For all results, detailed chart review revealed main category of presenting symptoms, patient location at time of testing (either inpatient or outpatient), sex, age, whether cancer was present at the time of testing or later detected, and the specific results of the testing. The data are summarized by category of testing and specific autoantibodies.
© 2021 The Author(s

Value of the Data
• The data provided are of value as paraneoplastic autoantibody testing may be over-utilized, potentially leading to excess costs and downstream impact on patients. • Clinicians, other researchers, or personnel in clinical laboratories might find this data useful as a reference for comparison. • Our data set would serve as a starting point for researchers interested in future investigations investigating utilization of paraneoplastic autoantibody testing. • The data provide information on paraneoplastic autoantibody testing at an academic medical center over a decade. • The data provide information for 1,446 paraneoplastic autoantibody tests performed on 1,338 unique patients.

Data Description
Paraneoplastic syndromes are rare conditions associated with autoantibodies that are classically produced by malignancies but may also occur in the absence of any neoplasm [1][2][3] . Testing for paraneoplastic syndromes often involves panels of autoantibody assays [4][5][6][7] . The appropriate utilization of paraneoplastic autoantibody panels has been debated, with multiple studies showing relatively low yield of diagnostic testing and a high rate of false positives [8][9][10][11][12][13] . A primary factor may be testing in populations without appropriate clinical phenotype and context for paraneoplastic syndromes, providing opportunity for education to promote improved utilization of this area of laboratory testing [9 , 14-17] . There is thus opportunity for collaboration between pathology and clinical services that commonly order paraneoplastic and autoimmune encephalitis autoantibody testing in the design of algorithms and guidelines for testing [4] .
This retrospective analysis study includes data (including detailed chart review) on 1,446 samples originating from 1,338 unique patients who had paraneoplastic autoantibody testing ordered at an academic medical center. The laboratory testing data includes panels of serum paraneoplastic autoantibody panels offered by two reference laboratories in the United States (Mayo Clinic Laboratories, Rochester, MN; ARUP Laboratories, Salt Lake City, UT) and a CSF paraneoplastic autoantibody panel by Mayo Clinic Laboratories. ARUP Laboratories additionally offers stand-alone assays for N -methyl-D -aspartate (NMDA) receptor autoantibodies in either CSF or serum as well as separate assays for the P/Q-type voltage-gated calcium channel (VGCC) autoantibodies or voltage-gated potassium channel (VGKC) autoantibodies in serum. Table 1 shows the laboratory panels analyzed in the present study and their constituent antibody tests. The antibody types are divided into those directed towards intracellular antigens, neuronal surface antigens, and neuromuscular antigens. The paraneoplastic serum panel at Mayo Clinic Laboratories is more extensive than that offered at ARUP Laboratories. The Mayo paraneoplastic CSF panel overlaps somewhat with the Mayo serum panel but does not include some antibodies tested for by the serum panel and also has some antibodies not included in the serum panel. These do not represent all of the paraneoplastic testing available at ARUP Laboratories and Mayo Clinic Laboratories, but rather those ordered at our medical center. Ordering options for panels and standalone assays have changed over time. Table 2 summarizes the total number of tests and unique patients for the various panels or stand-alone laboratory tests analyzed in the present study. There was a total of 1446 paraneoplastic autoantibody tests performed on 1338 unique patients, with some patients receiving more than one type of test (e.g., both Mayo serum and CSF paraneoplastic panels). A footnote to Table 2 also notes a small number of patients who had indeterminate results on testing if the reference laboratory had that category as a possible result based on quantitative signal. Table 3 summarizes the specific antibodies detected in the present study, which testing included the positive antibody (either a panel or stand-alone test), and the number of malignancies found.
From the paraneoplastic serum panel from Mayo, autoantibodies to striational antibodies (38/483, 7.9%) and VGKC (20/483, 4.1%) were the most common detected. Fig. 1 summarizes a breakdown of testing for the Mayo serum paraneoplastic autoantibody panel, with results classified into four categories: (a) one or more positive results on the panel in a patient with known or later discovered malignancy, (b) negative results on the panel in a patient with known or later discovered malignancy, (c) one or more positive results on the panel for a patient without a known malignancy, and (d) negative results on the panel for a patient without a known malignancy. Fig. 1 A shows the data by absolute number of orders. Fig. 1 B depicts the data with a breakdown by percent within each category of presenting symptoms.
There was some overlap in ordering between the Mayo serum and CSF paraneoplastic panels. We identified 45 patients who had both panels ordered on the same day, with the following breakdown: 37 patients where both panels were negative, 7 patients where the Mayo serum panel had one or more positive antibodies, but the CSF panel was negative (see below for details), and 1 case where both panels showed positive Purkinje cell antibodies (titer of 1:640 in      There were no patients for which the Mayo serum and CSF paraneoplastic panels were ordered concurrently more than once. There were also no examples where the Mayo CSF paraneoplastic antibody panel had a positive while the serum paraneoplastic antibody panel was negative. It should also be noted that AChR binding and striated muscle antibody assays were not included in the Mayo CSF paraneoplastic panel during the retrospective analysis period ( Table 1 ). For the ARUP serum and CSF NMDA receptor antibody tests, there were 31 unique patients who had both tests performed on the same date of collection (one patient had this happen two times for a total of 32 occurrences). In all cases except one, both tests were negative. There was a single occurrence in a 58-year-old male with constitutional symptoms of a positive CSF NMDA receptor antibody titer of 1:20 (reference range < 1:1) but with serum NMDA antibody titers that were negative.
The raw data for the study are included in Supplementary files 1-7.

Experimental Design, Materials and Methods
Although the overall retrospective analysis period was December 1, 2008 through November 30, 2018, some of the tests in the present study became available for ordering at various years after 2008. The details for the Supplementary Files above indicate the first date at which the particular panel or stand-alone test was available at our medical center. All data was obtained from patient data in the electronic medical record from the University of Iowa Hospitals and Clinics (Iowa City, Iowa, United States). A reporting tool within the electronic medical record, known as Epic Reporting Workbench, was used to identify all tests in Table 2 performed in the retrospective timeframe. Only data from patients who had the paraneoplastic tests or test panels described in the present study performed at the University of Iowa Hospitals and Clinics were included; no data was obtained from diagnostic vendors or reference laboratory databases for any of the laboratory assays used for clinical testing. Detailed chart review was performed on all results, regardless of whether the particular test was positive or negative. For classification of presenting symptoms, we followed the categories utilized by Alabareen et al. (autonomic, bulbar, cognitive, constitutional, coordination, motor, sensory, or vision) in their retrospective analysis of patients tested by the Mayo paraneoplastic autoantibody panel [8] .

Ethics Statement
The analyses had approval by the University of Iowa Institutional Review Board (protocol # 201812702) as a retrospective project.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.