Our data indicate that HBA, although rarely, can interfere with T immunoassay measurement, leading to false-positive and/or false-negative results that could trigger unnecessary investigations, misdiagnosis and/or inappropriate treatments.

HBA are produced against red blood cell proteins of diverse animal species that arise naturally or in response to various external stimuli, such as viral or bacterial infections. They can bind non-specifically and non-competitively to other proteins, including antibodies used in laboratory tests. This may cause interference in diagnostic test results, compromising their accuracy. The exposure to certain animals and animal-derived products is a risk factor for producing HBA, but they can also occur during vaccination, antibody-targeted therapies, blood transfusion and autoimmune disease [6, 7]. In our patients only the oldest man had a known risk factor, because he lives with a dog.

The prevalence of HBA ranges from 0.2–3.7% in the general population, with higher rates in sick and hospitalized patients (ranging from 0.2–15%). However, some reports suggest a prevalence of up to 40% in the general population, which may present different exposition to a foreign antigen (for example, vaccination, blood transfusion, antibody-targeted therapies, exposure to animals, etc).

The interferences on immunoassay caused by HBA are more frequent with sandwich assays, but they may be observed with competitive ones, resulting in falsely elevated analyte levels [5]. Therefore, by removing the HBA from the sample, the true value could be obtained. HBA may also cause false depression of serum ACTH, cortisol and thyroglobulin causing diagnostic difficulties [8].

This phenomenon is quite frequent in certain hormone assays (such as TSH and thyroid hormones) and quite rare in others, such as T. Indeed, rare cases of falsely elevated TT levels due to HBA are reported in women and adolescent. Moreover, elevated TT levels on immunoassay due to heterophile interference have been described in a patient with metastatic prostate cancer, initially thought to be refractory to medical castration, who underwent bilateral orchiectomy [1, 7, 9,10,11].

In addition, falsely elevated TT levels have also been identified in on case of anti-testosterone antibody presence, in rare cases of gammopathies (due to IgG interference), and in patients taking biotin (due to the use of biotinylated reagents). TT levels may also be overestimated due to cross-reactivity with other steroid hormones or medications with similar structural activity (Table 1) [4, 5, 9,10,11,12,13,14,15,16,17,18,19].

In our work, HBA interference was detected only with CLIA method. In fact, TT levels were higher than expected, both in women and in men, as measured by CLIA, whereas it was in the normal range when measured by competitive CMIA method in the woman. As previously discussed, serum may contain interference factors that may cause false results depending on different immunoassay. In the present work, CMIA kit did not give a false result probably due to the presence of different amount of blocking reagents against HBA incorporated in the kits. Indeed, given the frequent presence of HBA in the population, in recent years, the main assay kits have been equipped with reagents capable of reducing this interference. HBA are highly heterogeneous and variable in their concentrations among individuals, therefore no blocking reagent can completely ensure protection against such interference [3, 5].

As expected, in our woman a reduction in TT levels was obtained after treating the sample with the HBA reagent ‘HBT Scantibodies’ confirming that the high level was due to assay interference. The data was subsequently confirmed by LC-MS/MS method.

By contrast, both men showed an increase in TT level after treatment with HBA reagent. Rare cases of increased analyte level, rather than decreased, are described in literature after treatment with ‘HBT Scantibodies’ [20]. A possible explanation for this phenomenon is that the use of HBA-blocking agent removes weakly reacting HBA (of the IgG class), while leaving highly reactive specific IgM class antibodies. These specific IgM antibodies may then bind more strongly to the analyte antibody, leading to increased interference in the assay [5]. These data indicate that when the measured value after treatment with HBA reagent differs from the initial value (whether it increases or decreases), further investigations are needed to verify the real values. Only LC-MS/MS method can definitively provide the correct interpretation of T level, by removing these analytical interferences [2, 7]. In fact, also in our man TT levels obtained by LC-MS/MS were consistent with clinical data.

It is known that TT immunoassay measurements can have limited accuracy, especially at low concentrations, and can be subject to interference, while LC-MS/MS method is sensitive and specific, but it is more expensive and less available, therefore excluding its routinary use in all clinical laboratories [2].

When the clinical picture is not supported by the laboratory data, physicians need to consider the potential presence of analytical interference, because misdiagnosis can lead to unnecessary procedures for patients and can limit the quality of life.

In the woman the diagnosis of laboratory interference was insidious because she had clinical manifestations of hyperandrogenism, which led to many diagnostic investigations to exclude any possible organic cause. In fact, the misdiagnosis of a possible ovarian causes of hyperandrogenism led us to suggest an ovariectomy, according with gynecologist. The final finding of normal TT levels indicated that the woman suffered from a common pattern of androgenetic alopecia, which is also frequent in women. Similarly, in literature is reported a case of bilateral orchiectomy performed in a man with prostate cancer due to false high T level [11].

In our work also the men undergone many diagnostic procedures to exclude organic causes of hypertestosteronemia, causing considerable stress and negative economic impact. LC-MS/MS demonstrated the true TT levels, which were normal in the youngest one and consistent with the clinical picture of subclinical hypergonadotropic hypogonadism still compensated in the oldest one.