ctDNA surfaces occult malignancy months before diagnosis in early-onset autoimmune disease
In patients with recent-onset systemic sclerosis and dermatomyositis, plasma circulating tumor DNA detected cancers that had not yet come to clinical attention, often many months before a diagnosis was made. The signal was strongest in a cohort that has long worried rheumatologists and oncologists alike: adults whose autoimmune disease may be entangled with an underlying neoplasm that is still too small, too hidden, or too biologically quiet to declare itself on routine evaluation.
The preprint followed 134 adults with newly diagnosed disease, including 69 with systemic sclerosis and 65 with dermatomyositis. Over an observation window averaging 18 to 24 months, six cancers emerged, three in each disease group. Plasma ctDNA was detectable in four of those six patients, and in three of those four cases the mutations found in blood later matched mutations identified in the corresponding primary tumors. A median lead time of 13 months separated ctDNA detection from the eventual clinical cancer diagnosis, with a range from 1.4 to 17 months.
That interval matters at the bedside. A blood-based signal appearing a year before a tumor is clinically recognized creates a window in which imaging, biopsy, and surveillance could be focused rather than scattershot.
Autoimmune cancer risk met a molecular signal in blood
Systemic sclerosis and dermatomyositis have both been proposed as clinical models of cancer immunoediting. Autoantibody-defined subsets carry different cancer risks, yet malignancy often remains invisible at presentation, especially when constitutional symptoms, muscle weakness, skin disease, weight change, and inflammatory markers can be attributed to the autoimmune syndrome itself. In that setting, a tumor can hide inside the diagnostic noise.
The study used a targeted next-generation sequencing panel covering commonly mutated cancer genes to search plasma for ctDNA. The logic is straightforward: if a tumor is shedding DNA into the circulation, even at low levels, a plasma assay may detect molecular evidence of cancer before standard clinical workup does. This is especially plausible in diseases linked to paraneoplastic immune activation, where the cancer may be biologically active enough to trigger autoimmune manifestations long before it becomes radiographically obvious.
Six patients developed cancer during follow-up, and ctDNA identified four of those cases. Two cancers were not detected by the assay, a reminder that ctDNA is not a universal detector and will miss some tumors, especially if shedding is limited or the assay does not include the relevant alteration. Still, the pattern was notable: ctDNA positivity translated into an 11.4-fold higher relative risk of malignancy than ctDNA negativity within the observation window.
Predictive value was strong for ruling out near-term cancer, weaker for ruling it in
The most clinically useful number in this dataset may be the negative predictive value. Among patients with negative ctDNA results, the negative predictive value reached 98% for cancer within the follow-up period. In practical terms, a negative test substantially lowered the chance that an occult malignancy would declare itself in the near term.
The positive predictive value was much lower, at 20%. A positive plasma test therefore did not mean cancer was present in every case, and it should not be treated as a stand-alone diagnosis. Some patients without a cancer diagnosis still had detectable ctDNA, which could reflect a false-positive result, an undetected neoplasm beyond the observation window, or biological complexity not yet fully resolved by the assay. For clinicians, the result is less a verdict than a warning signal: a patient with positive ctDNA belongs in a more intensive diagnostic lane.
One sentence captures the tension in the data: ctDNA was good at identifying who should worry, but not good enough to define the tumor on its own.
Matching plasma mutations to later tumor tissue strengthened the biologic case
The most persuasive element of the report is the concordance between plasma and tissue. In three of four ctDNA-positive cancer cases, the mutations detected in blood matched alterations later found in the primary tumor. That concordance makes random noise far less likely and supports the interpretation that the assay was capturing real tumor-derived genetic material rather than nonspecific background variation.
For systemic sclerosis and dermatomyositis, the biology may be especially relevant. Autoimmune symptoms can precede cancer diagnosis because the immune system may be reacting to an evolving tumor antigenscape before the lesion becomes clinically obvious. ctDNA offers a molecular foothold in that process, converting an imprecise suspicion into a measurable signal. The assay does not replace age-appropriate cancer screening, targeted imaging, or biopsy, but it may help decide which patients need escalation sooner and which may be monitored with less uncertainty.
Surveillance strategy will need prospective testing before practice changes
This preprint argues for larger interventional trials, not immediate routine adoption. The authors frame the next question clearly: whether ctDNA-guided surveillance improves the timeliness of cancer detection, refines imaging and biopsy strategies, and ultimately improves outcomes in patients with systemic sclerosis and dermatomyositis. Those endpoints matter because the utility of an early signal depends on what clinicians do with it. A test that simply generates more anxiety and more scans would not earn its place.
Several limitations remain embedded in the design. The cohort was relatively small, cancer events were few, and follow-up averaged only 18 to 24 months, which means late cancers could still emerge after the observation window. The assay was targeted rather than genome-wide, so cancers without a covered alteration could be missed. The positive predictive value also shows that a positive result cannot be interpreted as a diagnosis without confirmatory evaluation.
Even so, the study establishes a credible proof of concept. In a population already enriched for cancer concern, plasma ctDNA found most near-term malignancies and frequently did so long before the clinical diagnosis was made. For diseases in which cancer can masquerade as autoimmunity, that is a biologically coherent and clinically consequential advance, pending validation in larger peer-reviewed studies and prospective surveillance trials.