YKL-40 and Ovarian Cancer

Abstract & Commentary

Synopsis: YKL-40 may represent a novel marker for the detection of early stage ovarian cancer. YKL-40 levels in early stage patients may also predict disease recurrence and survival. The use of YKL-40 in detection of early stage ovarian cancer deserves further investigation.

Source: Dupont J, et al. J Clin Oncol. 2004;22: 3330-3339.

The quest to identify accurate serum biomarkers for ovarian cancer has been problematic owing to their insensitivity and poor specificity as a decision tool and technological challenges facing the newer proteomic algorithms. Addressing this problem, Dupont and colleagues evaluated a new biomarker, YKL-40, in the diagnosis and prognosis of ovarian cancer. The biomarker was compared against CA-125 and CA 15-3—two known markers of ovarian pathology. To do this, YKL-40 levels were evaluated among 42 patients diagnosed with stages I to IV ovarian, fallopian tube, and primary peritoneal cancer, and 8 patients with recurrent disease, as well as 46 normal controls, 19 patients without prior cancer seen in a high-risk ovarian screening clinic, 42 patients in the same clinic with a history of cancer, and 33 patients with benign gynecologic disorders. Serum values for YKL-40 and CA-125 were obtained for all cohorts. An abnormal value was established from normal controls as greater than or equal to 62 ng/mL. Among the 4 patient cohorts without ovarian cancer, YKL-40 values were not significantly different and were within the normal limits. Patients with ovarian cancer recorded significantly higher mean levels of all their biomarkers compared to normal controls.

However, YKL-40 values appeared to be stage-discriminant, that is, higher with advancing stage and tumor burden. In addition, YKL-40 values were abnormal among all histologies, and a value above 80 ng/mL was prognostic for poor outcome. Compared with CA-125, YKL-40 was appropriately abnormal significantly more often in cancer patients and normal significantly more often in normal controls and patients with benign disease. Dupont et al conclude that these characteristics, along with the test’s availability and technical facility, make YKL-40 and important ovarian cancer biomarker that should be evaluated in large prospective trials.

Comment by Robert L. Coleman, MD

Most clinicians are well aware of the ovarian cancer survival track record. However, patients with early stage ovarian cancer are most often cured of their malignancy—some by surgery alone. Finding these patients, heretofore, has been largely serendipitous; usually identified during exploration for a pelvic mass. Nonetheless, improving our ability to either identify these patients and those at risk for malignancy fuels the labor to develop an effective screening program. If the current 1 in 4 ratio of early to late stage patients could be improved to just one in 3 the median overall survival would increase to about 5 years. Imagine if the ratios were reversed entirely!

In the current study, Dupont et al introduce a new, readily available serum biomarker, YKL-40, in an attempt to improve upon the current best, CA-125. Little is known about this marker’s function but it appears to express in conditions of extracelluar matrix degeneration and angiogenesis. The third marker in this study, CA 15-3, although not widely used, has been recently shown to add significantly to a predictive model, along with CA125, in patients with pelvic masses. Given the insensitivity of CA-125 alone, most efforts in the line of biomarker development have turned to cocktails of markers in the hope of improving accuracy of diagnosis. Although most trials have shown some increase in predictive power with this strategy, the varied prevalence of cancer in the study cohorts may more strongly affect the predictive capacity than biomarker performance. In addition, with more tests comes more cost limiting the generalization of practice.

There are several noteworthy findings about YKL-40 that supports Dupont et al’s recommendation to study this marker more fully. First, this is one of the few biomarkers which demonstrates abnormal levels in both mucinous and non-mucinous neoplasms. Although CA 19-9, a gastrointestinal marker, has been evaluated in mucinous neoplasms it is usually done once the diagnosis is made limiting it as a preoperative tool. Second, nearly two-thirds of stage I/II ovarian cancers had elevation of the marker. Historically, approximately 50% of stage I/II ovarian cancers are CA-125 negative (in the current report it was 35%). Third, the marker was rarely positive in benign disease or among normal controls. This may have been aided by patient selection as those with inflammatory conditions were screened out in some of the normal control cohorts. Lastly, the marker appears to be prognostic on a number of levels: stage/tumor burden, survival, and predisposition for disease. With regard to this latter point, 3 high risk but without disease patients with falsely elevated YKL-40 values were subsequently diagnosed with cancer with the following 18 months. It is important to recall that an association between false-positive elevation and subsequent ovarian cancer diagnosis accompanied the early work with CA-125. Although, subsequent studies confirmed that CA-125 alone was too insensitive to act as a screening tool, further prospective evaluation with YKL-40 in much larger cohorts is warranted. In addition, studies to identify this protein’s function may aid in construction of targeted therapy.

Further Reading

1. Hogdall EV, et al. Oncol Rep. 2003;10:1535-1538.

2. Lu KH, et al. Clin Cancer Res. 2004;10:3291-300.

3. Schutter EM, et al. Am J Obstet Gynecol. 2002;187: 385-392.

Dr. Coleman is Professor, Department of Gynecologic Oncology, University of Texas Southwestern Medical Center, Dallas.