Amy Carpenter
June 2025—With few FDA-cleared or -approved methods for tumor marker testing in body fluids, it is the laboratory that’s responsible for the tests.
“A specimen arrives at your door, and you have to figure out what, if anything, you’re going to do,” said Jonathan Genzen, MD, PhD, MBA, professor in the Department of Pathology, University of Utah School of Medicine, and chief medical officer, ARUP Laboratories, in an ADLM session last year. (For purposes of this story, he updated his comments in May.)
Body fluid tumor marker testing serves a clinical purpose, but it is largely supportive, not diagnostic. His advice: Be cautious and attentive to how testing may be used.
There’s reason for that, and the CA 125 ovarian cancer biomarker is an example, said Dr. Genzen, chair of the CAP Clinical Chemistry Committee. It is normally expressed in epithelial lining in the eye, trachea, female reproductive tract, cervix, and abdominal mesothelium. CA 125 can also be elevated in serous fluid even in nonmalignant conditions. “So even though you think about this as a tumor marker in serum or plasma, it may normally be present in ascites fluid depending on the cause.”
CA 19-9 is a biomarker for pancreatic ductal adenocarcinoma; it’s also known as Sialyl Lewis-a. “The production of CA 19-9 depends on the Lewis a-b gene product,” Dr. Genzen said. “If you’re Lewis a-negative, b-negative, you can’t even make CA 19-9, so it is not a useful tumor marker in those patients.” About six percent of individuals of Caucasian ancestry and 22 percent of individuals of non-Caucasian ancestry are Lea-b-.
The Clinical and Laboratory Standards Institute C49: “Analysis of Body Fluids in Clinical Chemistry” document describes several specimen pretreatments that are common in body fluid testing and could be used to address viscosity, among them low-speed centrifugation, freeze-thaw cycles, dilution, and hyaluronidase. Dr. Genzen provided examples of “how things can get a little messy.” Freeze-thaw cycles are listed as a way to decrease specimen viscosity. “I suspect it could work, but it also lyses cells that are in the specimen and could release things that are cellular into the fluid and degrade or even increase the target you are trying to measure.” This could lead to an inaccurate result.
Dilution to decrease specimen viscosity is also not advantageous, he said, “because it impacts your measuring ranges and you could miss something at the lower end.” Hyaluronidase powder pretreatment to decrease viscosity works only if the increase in viscosity is due to hyaluronic acid, Dr. Genzen said. “So it may work on a joint fluid, but it’s probably not going to work particularly well on a pancreatic fluid.”
Not listed in the C49 document but a pretreatment his laboratory sometimes uses is ultracentrifugation/airfuge. As an alternative, filtration too can be helpful, he said.
“The caveat to all of these—and it’s true for tumor marker testing—is that pretreatment steps must be validated,” he said, to ensure something accidental isn’t going on and going unrecognized.
The purpose of many of these validations is to eliminate the possibility of matrix interference, Dr. Genzen said.
Body fluids are a different matrix and sometimes behave differently than serum or plasma. The protein, lipid, and electrolyte content can affect the physical properties of a fluid type. “This can impact your ability to accurately pipette, and therefore it can impact the ability to get an accurate sample volume in the analytical phase of testing,” Dr. Genzen said.
There can also be container surface binding by certain analytes, “and body fluids are no exception.”
CLIA regulations for performance standards also apply to body fluid testing (§493.1253 Standard: Establishment and verification of performance specifications). Generally, these assays are not FDA cleared or approved. “So you as a laboratorian are responsible for establishing the performance specification of that particular assay, and this applies to laboratories that modify FDA-cleared or -approved test systems, which is what we commonly do with body fluids. We use a serum or plasma test on a different fluid type; that’s a change in intended use.”
CAP checklist requirement COM.40620 Body Fluid Analysis spells out the laboratory’s responsibility for validating and verifying methods for body fluid analysis and establishing best practices for interpretation.
One requirement “applies directly to body fluid testing that the laboratory offers as a routine, orderable test.” If the test is performed routinely on the fluid, it says, there must be a written procedure. COM.40475 Method Validation and Verification Approval–Nonwaived Tests applies.
“However—and this is one of the greatest things in the CAP checklist from my perspective—it specifically calls out that method performance specifications for blood specimens may be used for body fluids if the laboratory can reasonably exclude the existence of matrix interferences,” Dr. Genzen said. “So it’s essentially saying if you do your matrix studies, and you exclude the possibility of matrix interference, you can also rely on the validation data that you have for blood. It also says you can point from the procedure manual to the literature or alternatively do an appropriate study yourself.”
COM.40620 further requires that the reference intervals be defined and reported with body fluid results, unless the concentration of the analyte is reported in comparison to its concentration in a contemporaneously collected blood specimen. “Sometimes it’s not possible to derive a true reference interval because it’s not a fluid collection that is present in healthy individuals,” Dr. Genzen added. The checklist requirement provides suggested language: “If the result is to be interpreted by comparison to the patient’s blood, serum, or plasma, such results must be accompanied by an appropriate comment such as, ‘The reference interval(s) and other method performance specifications are unavailable for this body fluid. Comparison of this result with the concentration in the blood, serum, or plasma is recommended.’”
Said Dr. Genzen, “The COM.40620 checklist item also points out that reference interval citations from the package insert or published literature may be used as well” (COM.40605 Reference Intervals).
When the laboratory receives a fluid type that has not been previously validated because it is rare, rarely ordered, and not in the test directory, it is considered a clinically unique specimen, rather than a routine, orderable test. “And as long as it’s cleared by the laboratory director or their designee who believes that testing is clinically appropriate,” Dr. Genzen said, “the checklist item provides a pathway for testing, but it has to be accompanied by a comment.” The requirement suggests: “The reference interval(s) and other method performance specifications have not been established for this body fluid. The test result must be integrated into the clinical context for interpretation.”
“So it’s basically describing a path to move forward for clinically unique and medically important specimens,” Dr. Genzen said.
For those who validate body fluids, Dr. Genzen recommends Quick Guide to Body Fluid Testing, second edition, published in 2023.
Validation examples can be found in the literature, Dr. Genzen noted, citing a study of the performance of CEA and CA 19-9 in identifying pleural effusions caused by specific malignancies (Hackbarth JS, et al. Clin Biochem. 2010;43[13–14]:1051–1055).
ARUP Laboratories reported two studies excluding body fluid matrix interferences—one for commonly requested chemistry analytes with specimens tested on a Roche Cobas 8000 system and another for alpha-fetoprotein in pericardial, peritoneal, and pleural fluids with specimens tested on the Beckman UniCel DxI 800 Access immunoassay system (Owen WE, et al. Clin Biochem. 2015;48[13–14]:911–914; Owen WE, et al. Clin Biochem. 2018;56:109–112).
No systematic matrix interference was found in the analyte/body fluid combinations investigated on the Cobas 8000. “We performed studies on many different fluid types and observed excellent recovery,” Dr. Genzen said.
Matrix interference with AFP testing was also not observed for pericardial, peritoneal, and pleural fluids using the Beckman Access AFP assay on the UniCel DxI 800. “Great spiked recovery, excellent mixed recovery, and great precision were observed,” he said.
“But you have to do the studies,” Dr. Genzen said. “If it works well in serum and plasma, there’s a reasonable possibility, if not greater than that, that it will probably turn out fine in your body fluid, but you have to validate to know that for sure.”
“It takes someone to do the work. That’s the key part.”
Dr. Genzen also noted some pretreatment caveats. He and colleagues reported on an ARUP study that evaluated the effectiveness of hyaluronidase and ultracentrifugation pretreatment for a variety of body fluids before clinical chemistry testing (La’ulu SL, et al. Lab Med. 2021;52[5]:469–476). “Hyaluronidase treatment looked fine for cerebrospinal and pancreatic fluids,” Dr. Genzen said. “But we did see significant decreases in tumor marker concentration with ultracentrifugation for CA 19-9 in cerebrospinal fluid, and then a pretty dramatic decrease for CA 19-9 and CEA in pancreatic fluid. I don’t have a great explanation for that other than that maybe it’s just spinning down that cellular debris and maybe some of those tumor markers are in, or on, the cells. But if you didn’t do the study you wouldn’t know that, and you could end up reporting an erroneous low result.” In ARUP’s standard operating procedures is information about whether hyaluronidase or ultracentrifugation is appropriate for certain assays.
Hyaluronidase use is also not always without concerns, he added. He and colleagues at ARUP have done studies on contamination of different hyaluronidase powders. The powders, derived from bovine and ovine tissue sources, originate from testes and can contain significant amounts of testosterone (Genzen JR, et al. Clin Chem. 2019;65[1]:67–73). “We’ve observed contamination of hyaluronidase powders with testosterone, which can give you an artificially elevated result if that was the analyte you were testing for. Again, validate.”
In preparing for his ADLM talk last year, Dr. Genzen said he found little in the literature at that time describing best practices with tumor marker testing in body fluids and no review articles.
He turned to three textbooks: Tietz Textbook of Laboratory Medicine, Contemporary Practice in Clinical Chemistry, and Henry’s Clinical Diagnosis and Management by Laboratory Methods. Each has chapters on body fluids and tumor markers, often referring to a common set of fluid types: pleural, pericardial, CSF, pancreatic and peritoneal fluids; urine; stool; and fine-needle aspirations, particularly for thyroid cancer and parathyroid adenomas.
[divider]
[divider]
Mention of disclaimers and caveats are common, Dr. Genzen said. For example, in Tietz, under pleural fluid, for CA 19-9 and CA 125, it says they “have not been shown to have sufficient diagnostic accuracy to warrant their measurement.” For peritoneal fluid: “Overall, there appears to be little added value from the measurement of traditional tumor markers in ascitic fluid compared to serum.”
He added, “The textbooks are generally not recommending whether you should or should not order them.”
Dr. Genzen said the CLSI C49 Analysis of Body Fluids in Clinical Chemistry is an “excellent resource on how to validate a body fluid.” An appendix also describes various fluid types and markers found in those fluids, he said, many of which are tumor markers.
The Food and Drug Administration, however, doesn’t recognize C49 as a consensus standard, Dr. Genzen pointed out. “So if you follow a C49 validation, it doesn’t mean the FDA will say, ‘This is exactly what we’d expect of a validation of this type.’”
The CLSI document, too, includes a cautionary note regarding tumor markers: “‘The utility of tumor markers for malignancy diagnosis in body fluids has been controversial due to conflicting findings of sensitivity and specificity.’ So again,” Dr. Genzen said, “they’re also not recommending whether clinicians should or should not order these.”
How then to use tumor marker tests in body fluids? As supportive information when clinically indicated, with caution, and after validation has excluded the possibility of matrix interference.
Amy Carpenter is CAP TODAY senior editor.