The Diagnostic Trap: When PSA Results Mislead – A Case Report

Prostate cancer (PC) is the most common cancer among men, and the fifth leading cause of cancer-related death, worldwide [1]. The different incidence of PC diagnoses between geographical areas depends on several factors, e.g., extent of prostate-specific antigen (PSA) testing, age of the population, and ethnicity. The PSA level is used at diagnosis for risk stratification, and PSA density level (PSAD), which is PSA divided by prostate volume, is used together with prebiopsy MRI as a predictor of clinically significant PC [2, 3].

PSA is a glycoprotein enzyme, secreted by the epithelial cells of the prostate gland, and its plasma levels typically rise in the presence of prostate cancer. However, benign conditions, such as prostatitis and benign prostatic hyperplasia can also increase PSA levels, complicating its use as a screening tool [4].

After PC diagnosis, monitoring PSA levels provides valuable information about disease severity and aids in assessing therapeutic response following both medical and surgical treatments. PSA is also a reliable marker for detecting tumor recurrence after radical prostatectomy [5]. However, in metastatic castration-resistant prostate cancer (mCRPC), a combination of clinical evaluation, imaging and PSA measurements is used for monitoring, and PSA alone is not considered sufficient to change treatment [6].

Given the high incidence of PCs, and the consequent large volume of PSA tests performed, diagnostic assay manufacturers face increasing pressure to deliver high-throughput methods for rapid and accurate PSA measurements. This has driven the development of automated immunochemical analysis platforms designed to meet these demands. However, despite advancements, immunochemical analyses remain inherently susceptible to interference from substances like heterophilic antibodies, which can result in falsely elevated or suppressed results [7]. While these interferences are relatively rare, with reported frequencies ranging from 0.05 to 4% in immunochemical assays [8, 9], the widespread use of immunochemical tests, including PSA assays, may increase the likelihood of their occurrence. The susceptibility to analytical interference varies across immunoassay platforms and laboratory equipment, and discrepancies may be discovered either incidentally or through deliberate cross-checking using different methods.

Efforts to standardize equipment across hospitals aim to improve analytical consistency and facilitate the use of common reference intervals. This may also decrease the likelihood of incidental discovery of interferences.

Therefore, laboratories must continue to collaborate with manufacturers to enhance analytical quality, while clinicians should remain vigilant and consult the laboratory when clinical evaluations do not align with analytical results. We here present a case with a 77-year old man, diagnosed with mCRPC, where significant discrepancies in PSA measurements were observed between different laboratories.

In 2011, a 77-year-old man was diagnosed with prostate adenocarcinoma with a Gleason score of 7 (4+3), a plasma PSA of 12 µg/L, and no evidence of bone metastases. Following diagnosis, the patient was placed under active surveillance, including PSA monitoring. After 6 months, elevated PSA levels (31.6 µg/L) were detected, leading to treatment with bicalutamide, an androgen receptor antagonist. A nadir PSA of 1.5 µg/L was measured in 2014.

In 2015, the patient developed bone metastases with a concurrent PSA of 20 µg/L. Treatment was adjusted to include pamorelin, a gonadotropin-releasing hormone analog, with a following nadir PSA of 13 µg/L.

By 2016, new bone metastases were detected, accompanied by a PSA increase to 20 µg/L. With serum testosterone below 0.3 nmol/L, the patient was classified as having mCRPC.

At age 82, the patient, maintaining a good performance status, began treatment with enzalutamide, an oral androgen receptor signaling inhibitor and denosumab, a monoclonal antibody against receptor activator of nuclear factor κ-B ligand (RANKL), for prevention of skeletal-related events, alongside continued pamorelin. All clinical follow-ups for mCRPC occurred at the Department of Oncology, Lillebaelt Hospital, Vejle, Denmark.

At these visits, treatment effect was evaluated through a combination of clinical assessments and plasma PSA measurements every 4 weeks, supplemented by imaging-based tumor assessments every 3 months. To facilitate easier blood sampling, routinely planned blood sampling for PSA tests was performed at Esbjerg and Grindsted Hospital (EGS), Denmark near the patient’s home. Here PSA measurements were performed using the Abbott Architect assay.

On two occasions, PSA levels were also measured at Lillebaelt Hospital, Vejle (SLB), when results were missing at the clinical visits. In 2016, PSA results from SLB (75 µg/L, Roche Cobas e602) were significantly higher than those from EGS (20 µg/L, Abbott Architect). This discrepancy was initially attributed to disease progression and was not investigated further. Treatment with enzalutamide was continued. In December 2017, another PSA result from SLB (101 µg/L, Roche Cobas e602) was highly elevated compared to that from the local laboratory at EGS (25 µg/L, Abbott Architect). However, due to the absence of clinical signs of progression, the clinician suspected analytical errors and consulted the laboratory.

Initially, a dilution series was conducted at SLB (Roche Cobas e602) to assess for potential analytical interference. The results showed no signs of falsely elevated values, as the PSA levels remained consistent across dilutions. Unable to resolve the discrepancy with this approach, a sample was then sent to Aarhus University Hospital (AUH), using the Siemens ADVIA Centaur XPT assay, for further analysis. AUH (Siemens ADVIA Centaur XPT) found PSA levels similar to those measured at SLB (Roche Cobas e602) and likewise detected no signs of falsely elevated values through their dilution series. Subsequently, a dilution series at EGS (Abbott Architect) revealed an increase in PSA levels with each dilution, indicating a blocking factor causing falsely low PSA results. This is demonstrated in Table 1 and Figure 1.

It was concluded that the Abbott Architect PSA assay gave falsely low results in this patient due to an unspecified blocking factor. The patient was already diagnosed with mCRPC and these findings did not necessitate a change in treatment. At the following clinical control visits, PSA measurements were performed using the Roche Cobas e602 PSA assay.

This case highlights a scenario in which a patient with mCRPC had PSA levels measured at two different hospitals using different PSA assays, resulting in significant discrepancies. Initially, the higher PSA levels from the Roche Cobas e602 assay were suspected to be falsely elevated. However, further investigations revealed a blocking factor that when analyzing with the Abbott Architect assay led to falsely low PSA values.

Immunochemical assays are widely used in clinical laboratories due to their high specificity, sensitivity and fast turnaround time. However, inherent vulnerabilities for immunochemical assays relate to the antibody-dependent components, potentially resulting in analytical interferences [10].

These interferences may be caused by various factors, such as heterophilic antibodies, antibodies against detection systems (e.g., anti-ruthenium), M-components/paraproteins, or antibodies directed against the sought biomarker (e.g., anti-PSA antibodies). Such interferences can result in falsely high or falsely low results [10‒12]. Falsely elevated results have been more frequently documented in the literature, spanning a diverse set of analytical components, e.g., thyroid stimulating hormone and troponin I [13], estradiol [14], and PSA [13, 15, 16], resulting in either unnecessary radiologic imaging, medical treatments, or biopsies.

In comparison, falsely low results are less frequently described but can have clinical consequences when they occur. For example, a case report, by Loudas et al. [11], involved a 63-year-old man who, following prostatectomy, had falsely low to undetectable PSA levels, causing a delay in appropriate diagnosis and treatment.

Detecting these diagnostic pitfalls is critical, though challenging, as routine systematic investigations of interferences for all incoming blood tests are impossible in modern laboratories. However, once clinical suspicion is raised, methods, such as excluding pre-analytical errors, repeat testing, method comparisons, dilution tests, polyethylene glycol precipitations, and neutralization of interfering substances, can help identify interferences [10].

This case adds to the limited evidence of falsely low PSA results due to assay interference, emphasizing the need for clinicians to be alert and collaborate with laboratory personnel, when discrepancies arise between clinical assessments and laboratory results. In such situations, timely identification of interfering agents may change treatment decisions or prevent unnecessary diagnostic procedures. If considered relevant, the mentioned methods will be tools to identify such interfering agents, in order to avoid unnecessary diagnostic procedures or changes in treatment. In this case, the detection of discrepant PSA results from different methods was fortuitous.

Choosing between the different methods, for the detection of possible interfering substances, is an interplay between relevant clinical patient data, e.g., former immunization status or the presence of autoimmune diseases, influencing the occurrence of anti-animal autoantibodies, heterophilic antibodies and autoantibodies, as well as certain medical properties, e.g., biotin intake, and lastly utilizing the biochemical differences between the different assays [10]. In this case, several causes for interference in the Abbott Architect assay were theorized; heterophilic antibodies, a partial blockage of the assay capture antibody, autoantibodies with low affinity against PSA or even a defective PSA molecule. The investigation was unfortunately not finalized, but would have shed light on this rare occurrence.

As efforts to standardize laboratory equipment across hospitals continue, the likelihood of incidental findings of assay discrepancies may happen less frequently. This makes clinician’s vigilance in recognizing cases where laboratory results may be misleading even more critical.

We would like to thank the laboratory biochemists Erik Dalsgaard Lund and Inge-Lise Søtang Jacobsen for their excellent work.

Ethical approval is not required for this study in accordance with local or national guidelines. Written informed consent was obtained from the patient for publication of this case report. The CARE Checklist has been completed by the authors for this case report, attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000543661).

During collection of data for the case-report, Tobias Kromann-Tofting was employed at Department of Biochemistry and Immunology, Lillebaelt Hospital, Denmark. At the time of manuscript submission, Tobias Kromann-Tofting is employed at Siemens Healthcare A/S, Siemens Healthineers, Ballerup, Denmark. No other authors declare any conflicts of interest.

This study was not supported by any sponsor or funder.

Conceptualization: C.V.M., J.S.M., and P.D.S. Methodology, validation, and formal analysis: T.K.-T., J.S.M., and P.D.S. Investigation: T.K.-T. Resources: J.S.M. Writing – original draft: J.B.B.-K. Writing – review and editing: J.B.B.-K., C.V.M., T.K.-T., J.E.M., J.S.M., and P.D.S. Visualization: J.E.M. Supervision and project administration: J.S.M. and P.D.S. All authors have participated to the writing of the manuscript and approved its submission.

All data analyzed during this study are included in this article and its online supplementary material. Any additional inquiries can be directed to the corresponding author.