An Ovarian Mass in a Young Woman: Germ Cell Tumours and Their Markers

This post walks the category that owns young ovaries: the subtypes, the three markers that separate them, why CA-125 misleads here, and why the first operation is planned around fertility. An ovarian mass earns imaging and a workup. In this age group, most of them are not cancer.

Why age is the first thing that changes the answer

Germ cell tumours come from the cells that would have become eggs. That origin is why they behave nothing like the ovarian cancer most people have heard of, and why they arrive decades earlier. The NCI describes them as uncommon but aggressive tumours, seen most often in young women and adolescent girls, and notes they are frequently unilateral, meaning one ovary rather than both.

The age concentration is not a small effect. In patients under 20, up to 60% of ovarian tumours are of germ cell origin (Int J Mol Sci 2025).

Two things keep this in proportion. Malignant germ cell tumours of the ovary make up only about 2.6% of all ovarian malignancies (Diagnostics 2022). And within the germ cell family itself, roughly 95% are benign, almost all of them mature cystic teratomas (Diagnostics 2022). So this is a common category made mostly of benign things, with a malignant minority that lands in the decades where epithelial ovarian cancer is rare.

The subtypes behind the words germ cell tumour

The NCI's classification, which follows the 1973 WHO histological typing, lists dysgerminoma, endodermal sinus tumour, embryonal carcinoma, polyembryoma, choriocarcinoma, teratoma in immature and mature forms, the monodermal teratomas such as struma ovarii and carcinoid, and mixed forms. Those names carry real differences, and the pathology report is where the difference is settled.

Dysgerminoma is the most common malignant germ cell tumour of the ovary, around a third of malignant cases (Diagnostics 2022); the most common germ cell tumour overall is the benign mature cystic teratoma. Dysgerminoma is the female counterpart of testicular seminoma (Front Oncol 2014), and the marker-led workup rhymes with the one on the other side of the aisle, covered in our piece on a testicular lump in a young man. It concentrates hard in the young: 80 to 85% of patients are under 30 (Diagnostics 2022).

Yolk sac tumour, which the NCI still lists under the older name endodermal sinus tumour, is the third most frequent malignant subtype; its peak incidence is around age 30 and its median age 18 (Diagnostics 2022). The NCI calls these particularly aggressive, which is why the distinction from a dysgerminoma is not academic.

Immature teratoma is graded, and the grade does work. The current WHO classification, the 5th edition from 2020, grades it in two tiers: low grade, which was the old grade 1, and high grade, which was the old grades 2 and 3 (eClinicalMedicine 2024). Those older grades came from counting how many low-power microscope fields on a slide hold immature neuroectodermal tissue: grade 1 is less than one field, grade 2 is one to three, grade 3 is more than three (Cancers 2025). The NCI says survival after surgery alone is related to grade, and the figures behind that come from a series treated before the chemotherapy era.

Mature cystic teratoma is the benign counterpart and the common one, with its highest incidence between ages 20 and 40 (Diagnostics 2022). It is the reason most ovarian masses in this age group are not a cancer story at all.

The markers that answer the question, and the one that misleads

Three markers do the separating. The RCOG recommends LDH, alpha-fetoprotein and human chorionic gonadotropin in women under 40, while the National Academy of Clinical Biochemistry recommends considering only AFP and hCG, the two it counts as extensively studied and validated (Diagnostics 2022). The pattern across them points at the subtype. In that paper's table, dysgerminoma raises LDH with AFP normal and hCG variable; yolk sac tumour raises AFP with hCG normal; immature teratoma leaves hCG normal and may or may not raise AFP; choriocarcinoma raises hCG; and a mixed tumour can raise all three.

The NCI advises obtaining serum alpha-fetoprotein and human chorionic gonadotropin at diagnosis, and notes that these markers are not required for formal staging but that their persistence in the serum after surgery indicates unresected tumour. That second half is the part that gets underused: the markers earn their keep twice, once pointing at what the mass is and again saying whether the operation got all of it.

Then there is CA-125, the marker most people arrive already knowing, and the one most likely to mislead here. MedlinePlus describes the CA-125 test as mainly used to see whether ovarian cancer treatment is working or whether the cancer has come back. It lists endometriosis, pelvic inflammatory disease, uterine fibroids and liver disease as non-cancer causes of a high result, and adds that natural bodily functions such as a menstrual period and pregnancy can also raise CA-125. Every one of those is more common in a 22-year-old than epithelial ovarian cancer is, and a normal CA-125 proves very little about a germ cell tumour, because CA-125 is not that family's marker.

MedlinePlus is explicit that only a biopsy can diagnose ovarian cancer. hCG is the cleanest case: pregnancy raises it, which is why pregnancy is excluded before a raised hCG is read as coming from a tumour.

What the scan settles and what it leaves open

Imaging comes before any of this is named. The ACR Appropriateness Criteria for a clinically suspected adnexal mass rate transvaginal pelvic ultrasound as usually appropriate for initial imaging, rate CT as usually not appropriate at that first step, and describe MRI with and without contrast as a useful complementary study that helps characterise indeterminate lesions (ACR, 2023 revision).

What the ultrasound produces is a description, and it carries most of the early signal: size, whether the mass is cystic or solid or both, fat and calcification inside it, wall and septation detail, and blood flow. Those features are what separate a dermoid from something that needs a marker panel the same week, and our companion piece on telling an ovarian cyst from a cancer on imaging works through how those descriptors are weighed. For staging once a tumour is confirmed, the NCI lists CT, PET, MRI and transvaginal ultrasound among the tests used.

What imaging does not settle is the histology. The subtype, and in an immature teratoma the grade, come from tissue, which is the gap the markers only partly bridge beforehand.

Why the first operation is planned around fertility

The surgical plan in this group is written with future pregnancy in it from the first conversation, not as an afterthought. For stage I disease the NCI states that unilateral salpingo-oophorectomy, conserving the uterus and the opposite ovary, is accepted treatment for younger patients, and that it should be performed when fertility is to be preserved. One review goes further and calls fertility-sparing surgery the actual standard of surgical treatment for young patients with early-stage disease (Diagnostics 2022).

That is only possible because these tumours respond to chemotherapy. The regimen is bleomycin, etoposide and cisplatin, and the NCI states that BEP can cure the majority of patients even in stage IV dysgerminoma, and that with platinum-based combination chemotherapy the prognosis for the non-dysgerminoma subtypes has improved dramatically. These tumours carry a better prognosis than their epithelial counterparts (Front Oncol 2014), and the NCI's own summary is that they are generally curable if found and treated early. Exact survival varies by subtype, stage and grade.

How Healz reads an ovarian mass against the patient's age

An ultrasound report and a marker panel mean something only together, and only once the patient's age is part of the reading. A solid 8 cm adnexal mass with a raised LDH in a 19-year-old and the same two lines in a 62-year-old point in different directions, and the general reference range printed beside the number knows nothing about which one it is looking at. Healz is equipped with root-cause technology: it does not stop at the first label, it asks the questions a rushed visit skips, such as which markers were never drawn, and it cross-checks the case against more than a million rare cases to work out which subtype the pattern actually fits.

Everything sits in one chat rather than ten apps and four portals. Frontier AI works the whole file, so the ultrasound description, the AFP, the hCG, the LDH and the pathology report are read against each other instead of one page at a time. As a blood test ai analyzer, Healz is equipped to read the pathology report too, pulling the grade out of an immature teratoma and keeping it beside the marker trend. Memory pulls the medical facts out of every upload, keeps the abnormal values up front, dates each finding and connects them across months, so a marker that is falling after surgery looks different from one that has stalled. For the moment when you want a human looking at the same case, you can bring a board-certified doctor into the same chat for a second opinion.