Plasma Sarcosine Measured by Gas Chromatography-Mass Spectrometry Distinguishes Prostatic Intraepithelial Neoplasia and Prostate Cancer From Benign Prostate Hyperplasia

Pavel A. Markin, MS, PhD(c); Alex Brito, MS, PhD; Natalia Moskaleva, MS, PhD; Miguel Fodor, MD; Ekaterina V. Lartsova, MD; Yevgeny V. Shpot, MD, PhD; Yulia V. Lerner, MD; Vasily Y. Mikhajlov, MD, PhD; Natalia V. Potoldykova, MD; Dimitry V. Enikeev, MD, PhD; Alexey V. Lyundup, PhD; Svetlana A. Appolonova, MS, PhD


Lab Med. 2020;51(6):566-573. 

In This Article


In our study, based on predefined groups with diagnoses confirmed by biopsy, we found that plasma sarcosine had "good" and "very good" discriminative capability to contrast the noncancerous control (BPH) group versus the PIN group (AUC, 0.833) and versus the PCa group (AUC, 0.833). Using PSA and sarcosine together improved the overall diagnostic accuracy to detect both PIN and PCa compared to using these as single biomarkers.

Sreekumar et al found that metastatic specimens had increased levels of sarcosine in 79% of the analyzed tissue specimens.[8] Importantly, in their study, none of the noncancerous tissues had detectable levels of sarcosine.[8] Sreekumar et al found significantly higher sarcosine urine levels in biopsy-positive PCa patients than in the biopsy-negative controls.[8] Lucarelli et al found that serum sarcosine concentrations were higher in a group of patients with PCa than in patients with no evidence of malignancy.[9] Serum sarcosine showed a higher predictive value (AUC, 0.668) than did total PSA (AUC, 0.535) in patients with a PSA of <4 ng/mL.[9] In our study, contrasting PCa versus the control group, the AUCs for plasma PSA and plasma sarcosine were 0.854 and 0.833, respectively, regardless of PSA concentration. McDunn et al found elevations in Gleason grade 8 tumors versus benign tissues.[10] In our study, we performed subgroup analyses comparing different grades of Gleason in spite of being restricted to small specimen sizes. Plasma sarcosine did not differentiate individuals with PCa having a score of ≥7 versus <7 compared to plasma PSA.

In our study, plasma PSA had a high capacity to detect true-positive patients with PCa (high sensitivity, 92.6%); however, the test presented a low probability of detecting true-negative individuals (low specificity, 63.9%). In clinical practice, this low specificity can increase referrals of patients to unnecessary invasive procedures, such as biopsies, for further diagnostic confirmation. In clinical practice, based on our study population, if both markers are abnormal, there will be a high chance of having PCa, improving early detection of the disease. On the other hand, if any of the 2 markers are not abnormal, there will be a low chance of having the disease (high specificity, 91.7%). Interestingly, plasma sarcosine has 1 more feature. If plasma PSA is not abnormal, indicative of the absence of oncological alterations, but plasma sarcosine is abnormal, there will be a chance of detecting PIN as an early development of PCa. Classification of PCa versus BPH using plasma PSA alone and plasma sarcosine alone presented higher NPVs than PPVs. Using plasma PSA and plasma sarcosine together presented good PPVs and NPVs, with both values above 80%. However, it is important to note that these calculations are affected by the prevalence of having PCa versus the control group found in our study population. This prevalence was 31%. In populations where it is expected to find a higher prevalence, the PPV should increase while the NPV should decrease, and vice versa.[29]

Overall diagnostic accuracy to detect PIN or PCa versus BPH was highly impacted by using the 2 markers together. This overall parameter has to be interpreted with caution because it is influenced by prevalence, and the rest of the parameters should be weighted to make an overall interpretation. In our study, using plasma PSA and plasma sarcosine together presented an improvement in overall diagnostic accuracy accompanied by good diagnostic performance for the rest of the parameters.

The present study measured plasma sarcosine in order to contribute to the current knowledge regarding the use of sarcosine as a biomarker in the diagnosis of PCa. The study design was based on 3 well-discriminated groups with diagnosis made using biopsy specimens. A more informative study would have a narrower age range and collect more information, such as patients' race, family history, genetic profiling, and risk factors causing or predisposing them to develop PIN or PCa. We acknowledge that inference of causality is limited due to comparing groups in an experimental design without follow-up, such as a case-control study, a prospective cohort, or a randomized controlled trial. Statistically, it is possible that there was not enough power to detect significant differences due to limitations with the specimen sizes, especially when making comparisons after stratifying 2 groups with different Gleason scores in the PCa group. We did not include a control group with an absence of prostate gland enlargement. However, regardless of enlargement, our noncancerous control group comprised individuals diagnosed with BPH. They had a confirmed absence of oncological alterations by biopsy. More robust studies are needed to confirm the value of plasma sarcosine concentration as a biomarker and to establish the diagnostic impact of this marker on the therapies and the survival of the patients.