Confounding Hyponatremia and Hypochloremia in a Cancer Patient

Case Description

A 72-year-old female with a history of recurrent pancreatic ductal adenocarcinoma and a rising CA19-9 level presented to an outpatient clinic for a follow-up CT scan. She has locally recurrent and metastatic disease. The metastasis was diagnosed a few months prior by liver biopsy. She was previously enrolled into a clinical protocol, but developed a deep vein thrombosis (DVT) and a pulmonary embolism (PE) and thus was taken off protocol. While off protocol, her CA-19-9 levels had been rising. She presented in clinic for evaluation to restart traditional chemotherapy and for a follow-up abdominal computer topography (CT) scan. The patient received 150 cc of Omnipaque™ 300 prior to the CT scan. Laboratory tests ordered included a Complete Blood Count (CBC) and a Basic Metabolic Panel (BMP). Blood collections were performed 2 h after the CT scan. The samples were received in the laboratory, centrifuged, and analyzed on an Abbott C4000 analyzer. It was brought to the attention of the laboratory directory due to the very unusual nature of the gel formation on top (see sample) and the instrument error code. Figure 1, Panel 1shows a picture of the sample after centrifugation. The gel barrier on top prevented aspiration and resulted in an error code. The gel barrier was not completely formed, so the laboratory manager was able to manually insert a pipet around the gel and aspirate the sample. Abnormal laboratory results included: chloride = 82 mEq/L (Ref. Range: 98–109 mEq/L) and sodium = 111 mEq/L (Ref. Range: 133–143 mEq/L). These markedly abnormal values were not consistent with her clinical presentation or her previous laboratory values a month prior, which were in the normal range.

Fig. 1.

Panel 1: Original patient sample after centrifugation. Gel has moved to the top above the serum and blood. Panel 2: Replicated experiments showing an identical pattern as the patient sample. Sample A has 0.75 mL of Omnipaque contrast, Sample B has 1.5 mL, and Sample C has 3.0 mL.

Case Discussion

Laboratory errors are classified as preanalytical, analytical, and postanalytical. Many reports have confirmed that preanalytical errors represent the majority of error types in the clinical laboratory (1, 2). According to Hammerling (1), common causes of preanalytical errors include: inappropriate test request, order entry errors, misidentification of patient, inappropriate container, inadequate sample collection and transport, inadequate sample/anticoagulant volume ratio, insufficient sample volume, sorting and routing errors, and labeling errors. While most preanalytical errors occur outside of clinical laboratory control, it is important to monitor these factors as quality assurance indicators for the hospital or clinic. One specific source of preanalytical error is iatrogenic contamination of a specimen that is collected in a line, without proper flushing.

The sample in this case was collected from the same line used for radiocontrast infusion (Omnipaque™). As described by Sprititus et al. (3), the high-density molecules, such as those in radiocontrast, increase the density of plasma or serum, leading the gel in the separator tubes to partially or completely float to the top, which then can create a hard barrier between the gel and the serum. In addition, the gel formation on top will inhibit aspiration of the sample. On the Abbott Architect, sodium and chloride are measured by indirect ion-selective electrodes (ISE). This method uses the Nernst equation, as an electrical potential is created between membranes of a reference ion and the ion being measured (4). The iodine present in the radiocontrast can directly interfere with the sodium/chloride measurement (as can other ions in high concentration). The exact mechanism of the spuriously low sodium and chloride measurements is not identified and is beyond the scope of this case study. Possibilities include a dilutional effect or ion exclusion.

Finally, it should be noted that in addition to radiocontrast dye for CT studies, as shown by Proctor et al. (5), gadolinium infusions for MRI studies can also interfere with a number of clinical assays including angiotensin-converting enzyme, calcium, magnesium, iron, and zinc.

A follow-up BMP analysis several hours after the contrast infusion was drawn via a venipuncture showed a sodium value of 130 mEq/L (Ref. Range: 133–143 mEq/L) and chloride value of 96 mEq/L (Ref. Range: 98–109 mEq/L), consistent with the patient’s previous history. We decided to reproduce this patient’s scenario by spiking 10 mL tubes of fully drawn whole blood with 0.75 mL (Sample A), 1.5 mL (Sample B), and 3.0 mL (Sample C) of radiocontrast, respectively. Each 10 mL tube was a waste sample from patients with normal electrolyte values. The samples were spun and then visually inspected. Images of the samples are shown in Fig. 1, Panel 2. We analyzed the serum from these tubes on the Abbott Architect and the results are shown in Table 1.

Table 1.

Values from replicated experiment.

Sample name	Sodium value	Chloride value
Sample A	110 mEq/L	74 mEq/L
Sample B	<110 mEq/L	<68 mEq/L
Sample C	<110 mEq/L	<68 mEq/L

Note: The ‘<’ symbol for sodium and chloride, respectively, indicates that the result is beyond the lower limit of quantification of the instrument.

The results of the spiking experiment were consistent with the results obtained from the originally submitted specimen. As the concentration of the Omnipaque dye increased, the interference increased. In fact, samples B and C produced measurements beyond the limit of quantification (i.e., less than values for sodium and chloride). The image also demonstrated the formation of the gel barrier at the top of the tube. However, it should be noted that a dilutional effect could also impact the results and cannot be ruled out as a possibility for the low Na and CL values. Clearly, this large amount of radiocontrast dye can have a profound effect on the specimen. We believe clinicians should be aware of this issue as a source of interference in laboratory testing. Usually, the injected contrast will diffuse throughout the body, although it is still recommended that laboratory blood draws be done several hours after infusion. Even if it is drawn via a venipuncture, the radiocontrast can still cause an interference. In our case, it was confounded by the fact that the draw was from the same line used for infusion without proper flushing. This is an extreme example; however, this may occur quite commonly in a busy hospital setting if the clinical team is treating a patient in an emergent situation and needs radiology and laboratory studies quickly. It is important to stress that laboratory tests should be drawn before the radiocontrast infusion, and if a port or line is used, it should be flushed properly. Clinicians should interpret these values with caution, especially if the values do not match the clinical picture

Takeaways

• Indirect ISE measurements may be erroneous when there is contamination with radiocontrast dye.

• Radiocontrast dye can cause an abnormal separation after centrifugation and the hard barrier at the top of the specimen can prevent proper aspiration of the sample.

• Blood collections drawn by a line or port should be properly flushed and generally avoided, if possible.

• The Clinical Laboratory should report these errors to the clinical staff and monitor draw type as a quality assurance indicator.

• These types of preanalytical errors result in extra work for the laboratory staff, delay in care for the patient, and can possibly result in patient harm if the erroneous values are misinterpreted as accurate patient results.

Author Contributions

All authors confirmed they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved.

Authors’ Disclosures or Potential Conflicts of Interest

Upon manuscript submission, all authors completed the author disclosure form. Authors' disclosures and/or potential conflicts of interest: Employment or Leadership: None declared. Consultant or Advisory Role: None declared. Stock Ownership: None declared. Honoraria: None declared. Research Funding: This research was funded in part through the NIH/NCI Cancer Center Support Grant P30 CA008748. Expert Testimony: None declared. Patents: None declared.