Reading the Blood Instead of the Bone: ctDNA–Tissue Concordance in Metastatic Prostate Cancer

The clinical problem

Metastatic castration-resistant prostate cancer (mCRPC) is notoriously hard to biopsy. The disease loves bone, and bone is the one tissue that fights back against the needle: decalcification degrades nucleic acids, cores come back acellular or necrotic, and the patient endures a painful procedure that may yield nothing sequenceable. For a cancer type where somatic genotyping increasingly informs eligibility for targeted drug classes, that's a real bottleneck. If a tumor's molecular profile could be read from a tube of blood, it would sidestep the hardest part of the workup.

This is where plasma genotyping enters, and it raises a question that has to be answered before anyone trusts a blood-based result: how faithfully does circulating tumor DNA (ctDNA) reproduce the somatic alterations you'd find in matched tumor tissue? A blood test is only as useful as its concordance with the thing it's meant to replace.

The analyte and the assay

When tumor cells die, they shed fragmented DNA into the bloodstream. This ctDNA circulates mixed in with the much larger pool of cell-free DNA (cfDNA) released by normal cells—white blood cells especially. The analyte of interest, then, is a minority signal buried in a majority of background. The proportion of cfDNA that is tumor-derived is called the ctDNA fraction, and it is the single most important variable governing whether a plasma assay can see anything at all.

Plasma genotyping targeted sequencing of this cfDNA looks for two broad categories of alteration: point mutations and small indels, and copy-number alterations (CNAs) such as the gene amplifications and deletions that are common drivers in prostate cancer. The assay reports which alterations it detects in the plasma; the validation question is whether those detections match what independent tissue sequencing found in the same patient.

A result here establishes something narrow but important: the presence of a given somatic alteration in circulation at the time of the blood draw. What it cannot establish, when negative, is absence—and that asymmetry is the crux of interpreting these reports. More on that below.

The evidence

The anchoring study for prostate ctDNA–tissue concordance is Wyatt and colleagues, published in 2017, which compared plasma cfDNA against matched tumor tissue in a cohort of 45 patients with mCRPC [1]. The cohort size is worth naming up front: 45 matched pairs is enough to establish a direction and a magnitude, but it's small enough that I'd treat the specific percentages as working benchmarks rather than settled constants. They tell you what to expect; they don't tell you what will happen in your next hundred patients.

The study's central finding is best understood through the lens of ctDNA fraction. Among patients whose ctDNA fraction exceeded 2%, every somatic mutation identified in tissue was also detected in the matched plasma [1]. That is a striking result: when enough tumor DNA is actually present in the blood, the plasma reads back the tissue mutation profile completely. The 2% threshold isn't arbitrary—it marks the point below which the tumor signal gets too thin to reliably rise above background.

For copy-number alterations, the concordance between plasma and tissue was 88.9% [1]. CNAs are harder to call from plasma than point mutations because detecting an amplification or deletion depends on quantifying subtle shifts in sequencing depth across a region, and those shifts are proportionally diluted by non-tumor cfDNA. An 88.9% concordance for CNAs is therefore genuinely encouraging—it says that the categorical events that matter in prostate cancer survive the translation from tissue to blood reasonably well, at least when adequate tumor DNA is present.

Put the two findings together and the shape of the method becomes clear. Above the ctDNA fraction threshold, plasma genotyping is a high-fidelity mirror of tissue for mutations and a strong—though not perfect—reflection for copy number.

What a negative result does and doesn't mean

Here is the interpretive point that every reader of one of these reports should internalize. A positive plasma finding, in a sample with sufficient ctDNA, is informative and concordant with tissue [1]. A negative plasma finding is a different animal entirely. Because detection depends on ctDNA fraction, a sample with low tumor content may simply not carry enough signal to reveal an alteration that is genuinely present in the tumor. The absence of a detected alteration cannot be assumed to mean the absence of that alteration in the cancer—it may instead mean the blood draw caught the disease in a low-shedding state.

This is precisely why ctDNA fraction belongs on the report, not buried in a methods appendix. A clinician who sees a negative result without knowing the ctDNA fraction is reading an answer without knowing whether the test could have answered at all. A negative call from a sample above 2% ctDNA carries very different weight than a negative call from a sample where tumor DNA was undetectable.

Where it stands

The concordance data from Wyatt 2017 support a specific and limited claim: in mCRPC, plasma genotyping captures tissue-confirmed somatic mutations completely and copy-number alterations with high concordance, conditional on adequate ctDNA fraction [1]. That conditional clause is doing most of the work. The method's strength and its limitation are the same mechanism—dependence on the quantity of tumor DNA in circulation.

Several open questions follow. The 45-patient cohort establishes feasibility and sets the concordance expectations, but the durability of these figures across larger, more heterogeneous populations, across different assay platforms, and across disease states other than castration-resistant disease remains to be characterized within this evidence base. Readers should also note that the 2% threshold is a feature of the assay and cohort studied, not a universal physical constant; different assays may draw the line elsewhere.

The practical takeaway worth carrying from the report to the bedside is this: when a plasma genotyping result comes back negative, the first thing to look for isn't the gene list—it's the ctDNA fraction. If that fraction is high, a negative result is meaningful. If it's low or undetectable, the test hasn't told you the tumor lacks the alteration; it's told you the blood couldn't see it, and tissue remains the arbiter. Knowing which of those two situations you're in is the difference between a confident interpretation and a false reassurance.

References

1. Wyatt AW, et al. Concordance of circulating tumor DNA and matched metastatic tissue biopsy in metastatic castration-resistant prostate cancer. J Natl Cancer Inst. 2017. PMID 29206995.