In a Rare Ovarian Cancer, a Handful of Genes Drive Nearly Every Tumor Through the Same Signaling Pathway

A genomic atlas of 119 low-grade serous ovarian cancers reveals a disease dominated by missense mutations in the MAPK pathway — and a mutational landscape strikingly different from its more lethal high-grade counterpart.

Low-grade serous ovarian cancer is the quiet cousin of a far more famous killer. While high-grade serous ovarian cancer — the type that claims most lives — tends to strike fast, mutate wildly, and respond, at least initially, to platinum chemotherapy, the low-grade form does none of those things. It grows slowly, resists standard chemotherapy, and often recurs for years or decades. It is rare enough that many oncologists will treat only a few cases in a career, and poorly understood enough that clinical trials struggle to enroll.

What researchers at Memorial Sloan Kettering Cancer Center set out to map, using targeted sequencing of 119 tumors, was the molecular wiring of this disease — specifically, the mutations that hijack the MAPK signaling pathway, the cascade of molecular switches that tells a cell to grow and divide. The dashboard that follows lays out that map: which genes are mutated, how those mutations behave, where they cluster, and whether the data are clean enough to trust.

A Low-Mutation Disease, As Expected

The first question for any cancer cohort is how many mutations each tumor carries — its tumor mutational burden, or TMB. Cancers with high TMB, such as melanoma or lung cancer in smokers, tend to generate foreign-looking proteins that the immune system can learn to attack, making them candidates for immunotherapy. Low-grade serous ovarian cancer is not that kind of disease.

Across the 119 samples profiled here, the mutational burden clusters tightly at the low end of the spectrum, well below the 10 mutations per megabase threshold that oncologists often use to flag tumors for checkpoint-inhibitor therapy. The distribution is left-skewed, with the vast majority of tumors carrying only a handful of nonsynonymous mutations — changes to the DNA that alter the resulting protein. There are no dramatic right-tail outliers suggesting hypermutated cases contaminating the cohort. This is a clean, representative set of indolent serous tumors.

Missense Mutations Rule the Landscape

Not all mutations are created equal. A missense mutation swaps one amino acid for another in a protein, potentially changing its behavior. A nonsense mutation or frameshift inserts a premature stop sign, truncating the protein entirely. In cancers driven by oncogenes — genes that, when mutated, actively push a cell to grow — missense mutations dominate. In cancers driven by tumor suppressors — genes that normally restrain growth — truncating mutations are more common, because the goal is to destroy the brake.

Low-grade serous ovarian cancer is an oncogene-driven disease, and the mutational spectrum here confirms it. Missense variants tower over every other category. Nonsense and frameshift mutations exist but are comparatively rare. This is exactly what you would expect when the primary drivers are gain-of-function substitutions in KRAS, BRAF, and NRAS — single amino-acid changes at specific hotspot positions that lock these signaling proteins in their "on" state.

Low-grade serous ovarian cancer is an oncogene-driven disease, and the mutational spectrum here confirms it. Missense variants tower over every other category.

The Mutations Cluster on the Chromosomes You'd Expect

Because this is a targeted sequencing panel — not a survey of the entire genome — the chromosomal distribution of mutations reflects, in large part, where the panel was designed to look. And the panel was designed to look at MAPK pathway genes.

Chromosome 12, home to KRAS, and chromosome 7, home to BRAF, carry disproportionate shares of the total mutation count. Chromosome 1, where NRAS resides, also contributes. This pattern is not a surprise; it is a confirmation that the sequencing found what it was built to find. Other chromosomes show lower counts, consistent with secondary panel targets or off-pathway genes. What matters is that no single unexpected chromosome lights up with an anomalous spike that might suggest a recurrent hit outside the known pathway biology.

When variant types are broken down by chromosome, SNPs dominate uniformly — there is no chromosome where indels suddenly take over, which would have raised concerns about microsatellite instability or alignment artifacts in repetitive regions.

One Pathway, Four Genes, Mutual Exclusivity

The centerpiece of the dashboard is an oncoprint — a grid that lines up all 119 tumors in columns and the most frequently altered genes in rows, coloring each cell by the type of alteration present. It is, in effect, a portrait of the entire cohort's molecular identity at a glance.

The pattern is striking. KRAS, BRAF, NRAS, and to a lesser extent MET appear as the dominant players, with KRAS altered in the largest fraction of samples. Crucially, these alterations are largely mutually exclusive: a tumor that carries a KRAS mutation almost never carries a BRAF mutation as well, and vice versa. This makes biological sense. KRAS, BRAF, and NRAS are all nodes in the same signaling cascade. Activating one is sufficient to switch the pathway on; activating two would be redundant. Mutual exclusivity is the genomic fingerprint of a single-pathway disease.

Samples that appear as blank columns — tumors with no detectable alteration in any of the displayed genes — are not necessarily mutation-free. They may harbor gene fusions, such as in-frame BRAF rearrangements, which are known drivers of LGSOC but are invisible to a targeted mutation panel.

Hotspot Peaks and the Mutational Signature

Zooming in further, the downstream analysis maps mutations along the linear protein sequence, pooling all genes. The result is not a flat, featureless plain — it is dominated by sharp, narrow peaks at specific amino-acid positions. These are the hotspots: KRAS codons 12 and 13, BRAF codon 600, and their equivalents. Recurrence at the same residue across dozens of independent tumors is one of the strongest signals in cancer genomics that a mutation is a true driver, not a passenger along for the ride.

The underlying mutational signature reinforces the picture. When every base substitution in the cohort is tallied — reference nucleotide to alternate nucleotide — the dominant transition is C→T (equivalently G→A on the complementary strand). This is the hallmark of SBS1, the so-called clock-like signature associated with the spontaneous deamination of methylated cytosines that accumulates with age. It is the background hum of normal cellular aging, and it is the expected dominant process in a slow-growing cancer that lacks the dramatic mutagenic insults seen in smoking-related or UV-related tumors.

The Data Behind the Data: Quality Control

No genomic finding is more trustworthy than the sequencing that produced it. The final layer of the dashboard inspects the raw evidence behind every variant call: how many sequencing reads support the mutant allele, how many support the reference, and what the resulting variant allele fraction (VAF) looks like across the cohort.

The alt-allele read depth distribution shows that most called variants are supported by a comfortable number of alternate reads — well above the danger zone of fewer than five, where sequencing noise can masquerade as real mutations. The reference-allele depth distribution confirms high total coverage at mutated sites, as expected from a well-designed targeted panel. When alt and ref counts are plotted jointly, a dense cluster appears along the trajectory expected for heterozygous mutations in reasonably high-purity tumor samples — roughly 50% VAF. There is no alarming pile-up of low-confidence calls hugging the baseline, nor are there suspicious ultra-high-depth outliers that might indicate amplicon artifacts.

In short, the variant calls underlying this cohort's mutational portrait are well supported. The biology is not an artifact of noisy data.

The clinical implication of a single-pathway disease is both a constraint and an opportunity. Because virtually every low-grade serous ovarian tumor depends on the same MAPK signaling cascade — activated by a missense mutation in KRAS, BRAF, NRAS, or occasionally by a fusion event — the therapeutic target is narrow and well defined. MEK inhibitors, drugs that block the pathway one step downstream of these mutations, are already in clinical trials for this disease. But the same narrowness means that when the pathway evolves resistance, there may be nowhere else to strike. The blank columns in the oncoprint — the tumors without a detectable point mutation — are a reminder that the story is not yet complete. Gene fusions, epigenetic alterations, and co-occurring events in modulatory genes like NF1 all remain chapters to be written. For a cancer that grows slowly enough to give researchers time, these 119 genomes are a foundation, not a final word.