Correlation of CEA but not CA 19-9 as serum biomarkers of disease activity in a case of metastatic rectal adenocarcinoma

Abstract

We present the case of a 62-year-old-man with moderately differentiated adenocarcinoma of the rectum. This patient underwent neoadjuvant chemoradiation and surgical resection followed by adjuvant chemotherapy. After completing therapy, this patient had 2 instances of CEA elevation, both of which preceded the discovery of recurrent disease. While on treatment for these recurrences, CA 19-9 increased rapidly to 4,405. This CA 19-9 elevation persisted for approximately 4 months in the absence of clinical, radiographic or additional serologic evidence of progressive disease before returning to baseline. Shortly after this tumor marker normalized, a small area of locally recurrent disease was discovered. This case highlights the utility and pitfalls of colorectal cancer disease monitoring with CEA and CA 19-9. The differential diagnosis of CA 19-9 elevation is discussed in this report.

Keywords:

• CEA
• CA 19-9
• metastatic colorectal adenocarcinoma
• tumor marker

Introduction

Colorectal cancer (CRC) is among the most common malignancies in the developed world. Each year in the United States alone, more than 130,000 patients receive a new diagnosis of CRC.¹ This disease claims the lives of nearly 50,000 Americans annually, making it the third most common cause of cancer mortality in both men and women.²

Despite these somber statistics, considerable progress has been made toward improving outcomes through early diagnosis, effective treatment, and close surveillance of CRC. One such advance utilizes antigens that are commonly found in the peripheral circulation of patients with active malignancy, but are typically undetectable or less plentiful in healthy patients. Termed tumor markers, these substances are either shed directly by malignant cells or synthesized by normal cells in response to the presence of cancer. When considered in tandem with imaging, tumor markers provide clinicians with a minimally invasive means to approximate disease burden and to evaluate treatment response.

The tumor markers that are most often employed in the surveillance of CRC are carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9). The glycoprotein CEA is ordinarily produced during fetal development, when it participates in the formation of intercellular adhesions and signal transduction.³ These same functions are exploited in CRC to promote malignant cell survival and the formation of metastases.^4-6 In a clinical setting, CEA is best utilized as an indicator of active disease burden in patients with known CRC.⁷ Patients with elevated CEA at the time of diagnosis typically see this level normalize following surgery with curative intent. Persistent elevation of CEA after surgery suggests the presence of residual disease and portends a poor prognosis.^8,9 Similarly, repeat elevation of CEA in a patient with an initially positive serologic response to treatment is highly suggestive of recurrent disease^9-12 and often warrants further investigation if verified to be significantly elevated and/or consistently rising.

Carbohydrate antigen 19-9 (CA 19-9) is the second tumor marker that is often monitored in CRC. Bearing close resemblance to the Lewis antigen which studs the surface of red blood cells,¹³ this glycoprotein is a normal component of the glandular epithelium that lines the biliary tree, gastrointestinal tract, salivary glands, and other luminal structures.¹⁴ Similar to CEA, this antigen is often present within the peripheral circulation of patients with CRC and may indicate the extent of active disease. Unlike its counterpart, however, the role of CA19-9 in CRC surveillance is incompletely defined. Prior studies have established that elevated pre-treatment CA 19-9 is most common in patients with advanced disease^15-17 and independently predicts impaired survival regardless of disease stage.¹⁸ Despite this, research suggests that this tumor marker is of questionable value in the early detection of CRC recurrence¹⁹ and most clinical guidelines do not endorse routine monitoring of CA 19-9.^7,20,21 Nonetheless in clinical practice there are situations where CEA is not elevated while the CA 19-9 can be elevated as a marker of CRC disease activity.

The following patient we cared for further illustrates both the benefits and caveats inherent in utilizing CEA and CA 19-9 to monitor patients with known colorectal adenocarcinoma.

Clinical Case Report

A 62-year-old man was initially diagnosed with rectal cancer during a screening colonoscopy in 2008. The CEA level at the time of diagnosis was 2.2 ng/ml. The patient received neoadjuvant chemoradiation with capecitabine followed by low anterior resection of the primary tumor with clear margins. Surgical pathology revealed stage T3N0M0 moderately differentiated adenocarcinoma of the rectum with wild-type KRAS, BRAF, and NRAS.

Postoperatively, the patient received 12 cycles of FOLFOX (leucovorin, 5-FU, oxaliplatin) chemotherapy. Approximately 1½ years after completing therapy, the patient's CEA was noted to increase to 3 ng/ml. Less than 6 months later, a PET scan indicated a small area of FDG-avidity just lateral to the S2 vertebra. Further imaging revealed encasement of vasculature and the S2 nerve root by the presacral mass, rendering it effectively inoperable. The patient was treated with FOLFIRI plus bevacizumab for a total of 12 cycles with normalization of CEA and radiographic resolution of the presumed metastasis.

Shortly after completing therapy, the patient again experienced persistent CEA elevation to 113 ng/ml and transferred his care to our institution. PET-CT demonstrated progression of recurrent disease within the pelvis, but continued to show no evidence of extra-peritoneal metastasis. He was treated with stereotactic body radiotherapy (SBRT) followed by mFOLFIRI plus cetuximab combination chemotherapy. While on therapy, an alarming increase in CA 19-9 occurred without corresponding radiographic abnormalities or CEA elevation. This persisted for approximately 4 months before declining to baseline without specific intervention. PET/CT shortly after normalization of CA 19-9 uncovered a small focus of suspicious rectal thickening at the prior surgical site. Endoscopic biopsy was obtained and confirmed the presence of recurrent CRC.

This patient continued with at least cycle 13 of mFOLFIRI plus cetuximab at the time of writing this report with radiographically stable disease and good serologic response to treatment. He did not receive any additional locoregional therapy to the recurrent rectal tumor.

Discussion

This patient was diagnosed with colorectal adenocarcinoma during a colonoscopy at the age of 62. He received neoadjuvant chemoradiotherapy followed by surgical resection and adjuvant chemotherapy. After completing this treatment, the patient had 2 episodes of CEA elevation, both of which heralded the appearance of recurrent disease. During treatment for these recurrences, the patient also experienced a brief elevation of CA 19-9. The elevated CA 19-9 in this patient persisted in the absence of clinical, radiographic or additional serologic evidence of progressive disease for approximately 4 months before returning to baseline without specific intervention. Shortly thereafter, a small area of locally recurrent disease was discovered. A graphical representation of this patient's fluctuating tumor markers following the first discovery of recurrent disease, which occurred at an outside institution, is provided in Figure 1.

Figure 1. Relationship between radiographic findings and tumor markers CEA and CA 19-9 in a patient with metastatic colorectal adenocarcinoma.Legend: Patterns of CA 19-9 and CEA elevation in a patient with recurrent metastatic rectal cancer. CA 19-9 (carbohydrate antigen 19-9) Units/ml on the Y-axis above as indicated, CEA (carcinoembryonic antigen) ng/ml on the Y-axis below as indicated, FOLFIRI (leucovorin, 5-fluorouracil, irinotecan), SBRT (stereotactic body radiotherapy).

The relationship between CEA, CA 19-9, and disease activity described in this case is illustrative of both the value and caveats inherent in utilizing tumor markers to monitor patients with colorectal cancer. Twice during post-treatment surveillance, this patient demonstrated elevation of CEA and was subsequently found to have recurrent CRC – a pattern that is not uncommon in clinical practice. Among patients who experience normalization of CEA following treatment, a repeat elevation is both sensitive and specific for detecting recurrent disease.^19,22 This tumor marker is often the earliest indicator of recurrence, preceding clinical and radiographic evidence in 3-quarters of cases.^9-12 As a result, current ASCO guidelines recommend that CEA measurements be routinely obtained at 3 month intervals during postoperative surveillance and at 1–3 month intervals during systemic treatment for metastatic CRC.²³

The value of monitoring CA 19-9, on the other hand, is more controversial. While several studies indicate that its sensitivity and specificity rival that of CEA,^24,25 most suggest that CA 19-9 is vastly outperformed with respect to these measures.^19,26 Furthermore, CA 19-9 elevation trails radiographic detection of recurrent disease in 75% of instances, often by several months.^19,25,26 This tumor marker is also of questionable value in evaluating response to treatment. One study of nearly 1000 patients undergoing treatment for CRC found that decreasing CA 19-9 was concordant with radiographic response to treatment in only one-third of patients.²⁷ For these reasons, most clinical guidelines do not recommend the routine monitoring of CA 19-9 for CRC.^7,20,21

Despite its exclusion from guidelines, it is common practice at many institutions to trend CA 19-9 in addition to CEA. Consequentially, it is not uncommon for clinicians to face the conundrum of isolated CA 19-9 elevation in the absence of additional clinical, serologic, or radiographic evidence of recurrent disease. In this case, the picture was further clouded by the tumor marker returning to baseline without changing the patient's management. For these reasons, it is important to be cognizant of the sensitivity and specificity of this tumor marker, as well as the breadth of conditions that can cause its elevation.

Release of CA 19-9 into peripheral circulation occurs as a consequence of inflammation affecting glandular epithelium. This induces increased production of mucous glycoproteins such as CA 19-9, which are then extravasated into circulation.²⁸ As such, any condition that causes inflammation of the hepatobiliary system, gastrointestinal tract, or female reproductive system may increase the serum concentration of this tumor marker. Generally, nonmalignant and systemic inflammatory conditions such as hepatitis, gastroenteritis, diabetes, inflammatory myopathies, and idiopathic pulmonary fibrosis are associated with low-grade elevation of CA 19-9.^28,29 Scattered evidence also suggests that similar increases may be observed in chronic kidney disease and metabolic bone disease.³⁰ Greater increases, typically in excess of 1,000 U/mL, are generally observed only in cholestasis³¹ and cancers of the aforementioned glandular epithelia.³²

In this patient with known active malignancy and CA 19-9 elevation to 4,405 U/mL, the initial consideration was for progressive disease. Initially, PET/CT was obtained and revealed only stable disease with diminished FDG-uptake, consistent with response to treatment. Three months later, repeat PET/CT revealed a region of intensely FDG-avid rectal thickening that was later biopsy-proven to be recurrent CRC.

While this highly PET-avid tumor is ostensibly a satisfying explanation for this patient's transient CA 19-9 elevation, carefully evaluating the sequence of events casts doubt upon this theory. First, this patient's elevated CA 19-9 normalized rapidly and without any change in management. Had this tumor marker been increased secondary to progressive malignancy, such a decisive and spontaneous correction would be unlikely in the absence of additional treatment. Furthermore, this focus of recurrent disease was not visualized during the ascent of CA 19-9 but instead appeared while the tumor marker trended downward. This again seems unlikely, as the purported “effect” would have been resolving before the “cause” materialized.

Further review of the imaging suggests another possible explanation. Shortly before this patient's CA 19-9 elevation, he had complained intermittently of fatigue and nausea. While these symptoms could reasonably be attributed to his underlying malignancy or chemotherapy, serial CT scans revealed the presence of hepatic steatosis and cholelithiasis. In light of these findings, it is quite possible that this patient suffered transient steatohepatitis or cholestasis secondary to passage of a gallstone. As described earlier, such conditions are known to increase CA 19-9 to heights otherwise observed only in malignancy and would intuitively appear more consistent with the brief, self-limited episode described in this case.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.