Medical Policy


Subject: Gene Expression Profiling for Cancers of Unknown Primary Site
Document #: GENE.00018 Publish Date:    12/12/2018
Status: Reviewed Last Review Date:    11/08/2018


This document addresses the use of gene expression profiling (GEP) as a laboratory test to assist in the identification of the tissue of origin for cancers of unknown primary (CUP).

Position Statement

Investigational and Not Medically Necessary:

Gene expression profiling is considered investigational and not medically necessary as a technique to identify the tissue of origin for cancers of unknown primary site for all indications, including but not limited to:

  1. Determining the tissue of origin for a primary tumor; or
  2. Determining the tissue of origin for a metastatic tumor.

The National Comprehensive Cancer Network® (NCCN®) Clinical Practice Guideline (CPG) for occult primary tumors (V1.2019) defines CUP as “histologically proven metastatic malignant tumors whose primary site cannot be identified during pretreatment evaluation.” CUP are characteristically aggressive with a median survival of 6 to 9 months. In most individuals, chemotherapy is palliative and does not significantly improve long term survival. A primary tumor site is identified in fewer than 30% of individuals (Blaszyk, 2003). Diagnostic guidelines to aid in the identification of tumor types responsive to management have been developed and include a detailed history and physical, routine laboratory studies including immunohistochemistry (IHC) marker studies, and advanced imaging procedures. Several studies have suggested the feasibility of using molecular classification schemes with GEP to classify CUP (Greco, 2009; Tothill, 2005).

Notably, the NCCN CPG for CUP (V1.2019) states:

In an attempt to identify the tissue of origin, biopsy specimens are often analyzed by immunohistochemistry (IHC). In addition, gene expression profiling (GEP) assays have been developed to attempt to identify the tissue of origin in patients with occult primary cancers. Both methodologies offer a similar range of accuracy in tumor classification (approximately 75%). Thus far, the literature on GEP has focused far more on establishing a tissue of origin than on determining whether such identification leads to better outcomes in patients. Thus, while there may be a diagnostic benefit of GEP, a clinical benefit has not been demonstrated. Consequently, the panel does not currently recommend GEP for the identification of tissue of origin as standard management in the diagnostic workup of patients with CUP.

The NCCN CPG algorithm for the workup of CUP concludes:

Tumor sequencing and gene signature profiling for tissue of origin is not recommended for standard management at this time…There may be diagnostic benefit, though not necessarily clinical benefit. The use of gene signature profiling is a category 3 recommendation.

The NCCN defines a category 3 recommendation as, “Based upon any level of evidence, there is major NCCN disagreement that the intervention is appropriate.”

Tissue of Origin (TOO) GEP Test®

Microarray-based gene expression analysis, which captures data from tens of thousands of expressed genes in a single test, has been explored as a method to allow for more accurate classification of CUP, including those with high histological grade. The TOO Test (Cancer Genetics, Inc., Rutherford, NJ; formerly known as the Pathwork TOO Test and ResponseDX TOO, Response Genetics) was developed as a microarray-based gene expression diagnostic test that utilizes a panel of 1550 genes for determining the similarity of CUP to cancers from 1 of 15 known TOO. The results are reported as similarity scores, obtained by a statistical method involving pairwise comparisons by a machine learning algorithm. The technical performance (that is, reproducibility) and clinical validity (that is, sensitivity and specificity) of the test was analyzed on poorly and undifferentiated metastatic and primary tumors using fresh frozen tumor samples (Pathwork TOO Test, FDA, 2008) and formalin-fixed, paraffin-embedded (FFPE) tumor samples (Pathwork TOO Test Kit-FFPE, FDA, 2010) in the clinical validation studies submitted to the U.S. Food and Drug Administration (FDA) through the 510(k) process.

Dumar and colleagues (2008) assessed the performance characteristics of the Pathwork test in a cross-laboratory comparison study of 60 poorly and undifferentiated metastatic (77%) and primary (23%) tumors. Three academic and one commercial laboratory received archived frozen tissue specimens for procurement and processing at their individual sites. Steps performed by each of the four laboratories included tissue handling, ribonucleic acid (RNA) extraction, and microarray-based gene expression assays using standard microarray protocol. The resulting microarray data generated at each laboratory were sent in a blinded fashion to reference laboratory for generation of similarity scores for each type. Reports of the similarity scores were returned (blinded) to the pathologists at the four laboratories for their use in generating an interpretation. The physician-guided conclusion (PGC) showed an overall level of concordance between laboratories of 89.4% (range, 87.0% to 92.5%) in individual site comparisons. In terms of the reported results for all replicate specimens, the overall percent agreement of PGC with the known TOO (that is, clinical utility) was 86.7% (range, 84.9% to 89.3%) for all laboratories, with 4.6% discordant and 8.7% indeterminate results. Data were compared among the four laboratories to determine assay reproducibility. Correlation coefficients were between 0.95 to 0.97 for measurements of similarity scores; cross-laboratory comparisons showed an average 93.8% overall concordance between laboratories in terms of final tissue diagnosis. Limitations of this study include the small sample size of tumor tissues included in the analysis and the lack of reported confidence intervals. Confirmation of accurate and consistent results is needed from a larger sample obtained from a wide range of tissue samples at different stages of tumor progression.

The multicenter clinical validation study for the Pathwork test involved a comparison of the GEPs of 25 to 69 samples to each of the 15 known tumors on the Pathwork panel (average 36 specimens per known tumor). The specimens included poorly differentiated, undifferentiated, and metastatic tumors. A similarity score was given to 545 specimens and then compared to the available specimen diagnosis. Based on the results, the probability that a true TOO call was obtained when a similarity score of 30 or more was reported was 92.9% (95% confidence interval [CI], 90.3% to 95.0%), and the probability that a true negative tissue call was made when a similarity score of 5 or less was reported was 99.7% (95% CI, 99.6% to 99.8%). Overall, the Pathwork test performance comparing the profiles of the 545 specimens to the panel of 15 known tumor types showed a positive percent agreement with the reference diagnosis (sensitivity) of 89.4% (95% CI, 86.5% to 91.8%), negative percent agreement (specificity) of 99.6% (95% CI, 98.6% to 100%), non-agreement of 6.2% (95% CI, 4.4% to 8.6%), and indeterminate of 4.4% (95% CI, 2.8% to 6.5%) (FDA, 2008).

Monzon and colleagues (2009) conducted a multicenter blinded validation study of the Pathwork test that included specimens of poorly differentiated, undifferentiated, and metastatic tumors. Four institutions processed 547 frozen specimens representing no fewer than 25 specimens for each of the 15 tissues of origin on the panel. A similarity score was given to the specimens, which was then compared to the original pathology report that accompanied the specimen. The study found overall sensitivity of 87.8% (95% CI, 84.7% to 90.4%) and overall specificity of 99.4% (95% CI, 98.3% to 99.9%). Performance within the subgroup of metastatic tumors (n=258) was found to be slightly lower than that of the poorly differentiated and undifferentiated primary tumor subgroup, 84.5% and 90.7%, respectively (p=0.04). In an analysis of relevant subgroups, the rates of agreement between the test result and the reference diagnosis ranged from 94.1% for breast cancer specimens (n=68) to 72% for gastric and pancreatic cancer specimens (n=25 each). Performance differences between tissue sites were statistically significant (chi-squared = 42.02; p=0.04; degrees of freedom [df] = 28; n=547). Rates of agreement between the test results and the reference diagnosis were 88%, 84.4%, 92.3%, and 89.7% at each of four study sites, respectively; these differences were not statistically significant. A limitation of the study as acknowledged by the investigators was the inability to independently verify the reference diagnosis used to assess the accuracy of the test, as the diagnosis was extracted from the surgical pathology report that accompanied the specimen at the time it was banked. This lack of independent confirmation of the original pathology could result in an over- or underestimation of the test’s accuracy.

Dumur and colleagues (2011) attempted to verify the analytic and clinical performance of the Pathwork test on 43 poorly differentiated and undifferentiated tumor samples, including 6 off-panel cases and 7 CUP cases. The results reported a 97% (95% CI, 80.4% to99.8%) agreement between the Pathwork test result and the complete diagnosis, which included clinical correlations and IHC staining, after the original diagnosis. The investigators concluded that for off-panel and CUP samples, the tissue type and the cell type may be confounded by the Pathwork test; therefore, careful clinicopathologic assessment is needed when interpreting results from the test.

FFPE Testing

The multicenter clinical validation study for the Pathwork TOO-FFPE test (FDA, 2010) involved a comparison of the GEPs of 25 to 57 samples of poorly differentiated, undifferentiated, and metastatic tumors to each of the 15 known tumors on the Pathwork  panel (average 31 specimens per known tumor). A similarity score was given to 462 specimens and then compared to the available specimen diagnosis. Based on the results, the probability that a true TOO call was obtained when a similarity score was reported was 88.5% (95% CI, 85.3% to 91.3%). Any tissue type with a similarity score of five or less had a probability of not being the TOO, reported at 99.8% (95% CI, 99.7% to 99.9%). The overall performance of the Pathwork TOO-FFPE test when comparing the profiles of the 462 specimens to the panel of 15 known tumor types showed a sensitivity of 88.5% (95% CI, 85.3% to 91.3%), specificity of 99.1% (95% CI, 97.6% to 99.7%), non-agreement of 11.5% (95% CI, 8.7% to 14.7%).

Analytical performance characteristics of the Pathwork TOO-FFPE test were subsequently investigated in a cross-laboratory comparison study of 60 poorly and undifferentiated metastatic (45%) and primary (35%) tumors. Each of the 15 tumor tissue types were represented by four specimens each, with the exception of breast (n=3) and soft tissue sarcoma (n=5). Samples were distributed among three laboratories for procurement and processing at their individual sites; data collected was compared between the three laboratories to determine assay reproducibility. Correlation coefficients were between 0.92 to 0.93 for pathologists’ interpretations of the similarity scores, and cross-laboratory comparisons showed an average 82.1% overall concordance between laboratories in terms of final tissue diagnosis (FDA, 2010).

Grenert and colleagues (2011) sought to determine the performance of the Pathwork TOO-FFPE test on 45 blinded, archived clinical specimens from an independent university laboratory. Total RNA was processed to prepare labeled c-deoxyribonucleic acid (DNA) for hybridization to Pathchip microarrays. Hybridization data was analyzed with a 2000-gene classification model to quantify similarity between RNA expression of the study specimens and the 15 tissues on the test panel. A total of 37 of the 44 specimens successfully processed met the study inclusion criteria. Of these 37 cases, 35 cases (95%) gave results that were in agreement with the reference diagnosis. In no case was the reference diagnosis ruled out. The authors concluded the test gave a high agreement with the reference diagnosis and can be a useful tool in cancer diagnosis. Despite the inclusion of samples in the current study from all of the tissue types represented on the test panel, the sample size does not allow an estimate of the performance of the test on individual tissues, which included five metastatic tumors and six high-grade poorly differentiated primary tumors. The results of this evaluation were not in agreement with the reference diagnosis in 3 of the 11 individual tissue types (27%). This demonstrates a limitation of GEP, as profiles may overlap and application of the results are dependent on additional interpretation and correlation of the test results with other histopathologic, radiologic, and clinical findings.

Pillai and colleagues (2011) performed a validation study on the Pathwork TOO-FFPE test to determine the TOO in 462 metastatic, poorly differentiated, or undifferentiated FFPE tumor specimens, all of which had a masked reference diagnosis. The microarray data files were analyzed using a 2000-gene classification model. The algorithm, developed by the authors, quantified the similarity between RNA expression patterns of the study specimens and the 15 tissues on the test panel. Overall agreement of test results with the reference diagnosis was 88.5% (95% CI, 85.3% to 91.3%). The sensitivity was 88.5% and the specificity was 99.1%.The reproducibility study performed at three sites showed 89.3% (133 of 149 samples) concordance between the laboratories. The study however, did not discuss how these results would be used to improve outcomes in the clinical management of individuals with metastatic, poorly differentiated or undifferentiated tumors.

Stancel and colleagues (2012) performed a validation study using the Pathwork TOO-FFPE test in 27 metastasis-positive body fluid specimens from different body sites (for example, pleural, ascites, pericardial, and pelvic wash fluids) obtained from individuals with known diagnoses. Nine specimens from nonmalignant body fluids were included as controls; RNA was extracted from FFPE tissue. A total of 17 of 27 metastasis-positive samples were non-necrotic with ≥ 60% tumor and yielded sufficient RNA. Of these samples, 94.1% (16 of 17) cases were in agreement with the available diagnosis. Of the 9 negative control samples evaluated, 7 (77.8%) samples demonstrated microarray expression profiles most similar to lymphoma, consistent with the predominance of inflammatory cells in these specimens. The authors concluded these results demonstrated that FFPE cell blocks from cytologic body fluid specimens yield adequate diagnostic material for the Pathwork TOO-FFPE test and can be used in the workup of individuals with CUP. The results of this retrospective study are limited in drawing conclusions due to the small size of the sample and control groups.

Handorf and colleagues (2013) prospectively conducted a multicenter blinded validation study comparing the diagnostic accuracy of GEP with IHC in identifying the primary site of 157 FFPE specimens from metastatic tumors with known primaries, representing the 15 tissues on the Pathwork TOO test panel. Four pathologists rendered diagnoses by selecting from 84 stains in two rounds. Overall, the Pathwork TOO-FFPE test accurately identified 89% of specimens, compared with 83% accuracy using IHC (p=0.013). In the subset of 33 poorly differentiated and undifferentiated carcinomas, the Pathwork TOO-FFPE test accuracy exceeded that of IHC (91% to 71%; p=0.023). In specimens for which pathologists rendered their final diagnosis with a single round of stains, both IHC and the Pathwork TOO-FFPE test exceeded 90% accuracy;  however, when the diagnosis required a second round, IHC significantly underperformed the Pathwork TOO-FFPE test (67% to 83%; p<0.001). These findings confirmed the ability of IHC to diagnose the majority of metastatic tumors and suggested that GEP may be an alternative to IHC in the workup of occult primaries when the initial round of staining proves inconclusive. The clinical utility of the Pathwork TOO-FFPE test was not evaluated in this study along with tumor types that were off-panel. This limitation biased estimates of both the Pathwork TOO-FFPE test performance and the pathologists’ performance, as the evaluating pathologist also knew the tumors belonged to 1 of the 15 panel types.

Frozen (FRZ) Assay Testing

The predictive value of the Pathwork TOO-FRZ assay was evaluated on a diverse collection of malignancies classified into 1 of 4 groups: common morphology from a tissue type included in the TOO-FRZ assay (n=29), uncommon morphology from a tissue type included in the TOO-FRZ assay (n=10), tumor from a tissue type not included in the TOO-FRZ assay (n=3), and malignancies of unknown primary (n=7) (Beck, 2011). The authors reported a strong diagnostic performance for common morphologies from tissue types on the TOO-FRZ (overall accuracy, 26 of 29 cases [90%, 95% CI; 73% to 97%]), with “perfect performance” in all tissue types except gastric (0 of 2) and pancreatic (1 of 2) tissues. There was, however, a significant decline in performance for uncommon morphologies from tissue types included in the TOO-FRZ assay: 6 of 10, or 60% of cases with an indeterminate result; 1 of 10 or 10% of cases with an incorrect prediction; and 3 of 10 or 30% of cases with a correct prediction. For tumors from tissue types not included in the assay, there was an incorrect prediction in 2 of 3 cases. For the seven malignancies of unknown primary in the study set, the TOO-FRZ provided a likely clinically useful result in only 2 of 7 cases. Another limitation of the TOO-FRZ assay is that it does not clearly address the diagnostic subtypes within a tissue site of origin, as it was unable to differentiate between ovarian and uterine tissue origins, which is important in the clinical management of metastatic tumors from these sites.

CancerTYPE ID® Test

The CancerTYPE ID test (bioTheranostics Inc., San Diego, CA) is a quantitative real-time polymerase chain reaction (RT-PCR) assay that measures and integrates the expression of 92 genes, including 87 cancer-associated genes and five reference genes. The test RNA is extracted from FFPE tumor specimens, and complementary DNA (cDNA) is generated for the assay. The specimen’s expression profile is then compared to a database of tumor DNA expression data to distinguish 30 tumor types and 54 histological subtypes, reported as the probability of the occurrence of each cancer type. The CancerTYPE ID is the only assay test that includes identification of adrenal tumors. The results of an industry-sponsored study, based on analytical validation of the CancerTYPE ID reference database, reported the overall sensitivity and specificity of the test was 87% and > 99%, respectively, at the main tumor type level (30 classes) and 85% and > 99%, respectively, at the histologic subtype level. This data provides a preliminary estimate of classification performance. A separate test set of 187 FFPE tumor samples representing 28 of 30 main cancer types showed the 92-gene assay had an overall clinical accuracy of 83%, with 71% of the main tumor classes demonstrating a ≥ 80% sensitivity. In the observational arm of the study (n=300 cases), the 92-gene assay moderately confirmed the suspected primary tumor in 78% (43 of 55 cases) of those samples submitted with a single diagnosis. Within cases submitted with a differential diagnosis (n=133), the 92-gene assay was consistent with one of the suspected primary tumors in only 74% (n=99) of the cases. A limitation of this observational series is that the definitive reference diagnosis associated with each case was unknown. Considering the retrospective design of this study and the lack of identification of the actual TOO in most cases, the accuracy of the CancerTYPE ID assay test is not easily established. A prospective study is needed to assess whether the predictions from this assay will assist in the selection of treatment options that improve health outcomes for individuals with CUP (Erlander, 2011).

Kerr and colleagues (2012) conducted a prospective, blinded validation study of the CancerTYPE ID assay using tumor specimens from individuals evaluated at the Mayo Clinic, Massachusetts General Hospital, and the University of California-Los Angeles. Approximately 50% of the samples were metastatic tumors and the remaining samples were moderately to poorly differentiated primary tumors. Inclusion criteria included FFPE tumor block < 6 years old, 40% tumor material, diagnostic certainty, and minimal necrosis. A total of 957 tumor specimens were sent for processing by the CancerTYPE ID assay; however, quality control failures excluded 9 specimens and an additional 158 specimens were excluded per protocol. Of the 790 evaluable samples analyzed by the CancerTYPE ID assay, 43 (5.4%) were not classifiable by the assay. For the main 28 tumor types, the CancerTYPE ID assay had a sensitivity of 87% (95% CI, 84% to 89%) and a specificity of 99% to 100%. The lowest specificity was for endometrial cancer (43%; 95% CI, 26% to 70%). The positive predictive values (PPVs) for individual cancer types ranged from 61% (intestine) to 100% (multiple tumor types); negative predictive values (NPVs) were 99% or 100%. The PPV and NPV were not reported for the entire set of data. The sensitivity of the CancerTYPE ID assay did not vary significantly by cancer spread (metastatic versus primary), histologic grade (1, 2, or 3), or specimen type (limited biopsy versus excision). Brachtel and colleagues (2016) examined a subset of samples from 109 individuals with limited tissue from the Kerr study (2012) in addition to 644 other consecutive cytology samples. The sensitivity for tumor classification in the 109 samples was 91% (95% CI, 84% to 95%), or consistent with the larger sample. A sensitivity of 87% (95% CI, 84% to 89%) was estimated from the other 644 cases. A limitation of the study was that regardless of the high analytical success rate, “clinical validation by definitive follow-up could not be performed at this point.”

Hainsworth and colleagues (2012) examined the treatment and clinical course of 42 individuals with CUP that were predicted by the CancerTYPE ID assay to have tumors of colorectal origin. A total of 32 individuals received either first-line (n=24) or second-line (n=8) colorectal cancer treatment based on the CancerTYPE ID assay results. The overall response rate for individuals who received colorectal cancer treatment was 50% (20 of 40 cases), significantly higher than for individuals who received empirical treatment for CUP (17%, 3 of 18 cases). The median survival of individuals who received site-specific therapy for colorectal cancer was 27 months. The authors suggested that individuals predicted to have a colorectal site of origin by molecular tumor profiling had median survival when treated with site-specific regimens that was similar to survival in individuals with known metastatic colon cancer, and substantially better than the historical median survival for individuals with CUP (range 8-11 months) when treated with empirical CUP regimens. Additional study is needed in the form of a larger prospective trial to confirm these observations.

Hainsworth and colleagues (2013) conducted a multisite, prospective case series of the CancerTYPE ID assay to attempt to direct site-specific therapy for previously untreated individuals with CUP. The FFPE biopsy specimens for this study included adenocarcinoma, poorly differentiated adenocarcinoma, poorly differentiated carcinoma, or squamous carcinoma. Of the 289 participants enrolled, 252 had adequate biopsy tissue for the assay. The molecular profiling assay predicted a TOO in 247 (98%) of 252 participants. Sites most commonly predicted were biliary tract (18%), urothelium (11%), colorectal (10%), and non-small cell lung cancer (7%). A total of 119 assay predictions were made with ≥ 80% similarity score and the remaining were below 80% probability. A total of 29 participants did not complete the study due to decreasing performance, brain metastases, or by joint decision with their physician. Of the remaining 223 participants, 194 (87%) received assay-directed chemotherapy, and 29 received standard empiric therapy. The median survival time was 12.5 months for the 194 participants who received assay-directed chemotherapy, which was found to be within the a priori-specified improvement target of 30% compared with historical trial data on 396 performance-matched CUP individuals receiving standard empiric therapy at the same center. However, a clinical trial is needed where participants with CUP are randomized to receive treatment based on the results of these types of molecular profiling assay tests, or based on standard diagnostic procedures to determine the clinical utility of molecular profiling testing for CUP.

Greco and colleagues (2013) further evaluated the clinical validity of the CancerTYPE ID assay to accurately identify the TOO in individuals with CUP. The accuracy of the CancerTYPE ID assay performed on archived material was evaluated by comparison with latent primary tumor sites found months or years later, initial single diagnoses by IHC, and additional directed IHC and/or clinicopathologic findings evaluated after the CancerTYPE ID assay diagnoses. The diagnostic accuracy of the assay was supported by a high level of agreement with identified latent primaries (18 of 24 individuals, 75%) and single IHC diagnoses (40 of 52 individuals, 77%), and additional directed IHC and/or clinical/histologic findings (74%) prompted by the assay diagnoses. The CancerTYPE ID assay complemented standard pathologic evaluation in determining the TOO in individuals with CUP, particularly when IHC was inconclusive. This level of accuracy is similar to that already documented with other commercially available molecular profiling assays when tested on biopsies from individuals with known primary cancer. The authors cited limitations exist, in general, for all molecular profiling diagnostic assays including: 1) in a small fraction, no information is provided because of either technical failure (usually from inadequate biopsy) or an indeterminate result; 2) the assay diagnoses are not 100% accurate even when performed on known cancers (mean accuracy is about 85%); 3) overlapping gene expression of several neoplasms may cause incorrect diagnosis of the TOO; and 4) because molecular profiling assays depend on panels of known cancers for comparison with GEP of the unknown sample, several cancers not represented in the panels “off-panel” neoplasms may be incorrectly diagnosed.

Greco and colleagues (2015) retrospectively reported on use of the CancerTYPE ID assay on archived samples from 30 of 751 (4%) individuals with CUP and poorly differentiated neoplasms evaluated between 2000 and 2012 at Tennessee oncology and cancer care centers. A total of 28 of 30 individuals presented with advanced cancers and had no anatomical primary site detected after a standard work-up for CUP. A diagnosis was assigned by GEP in 25 of the 30 biopsy specimens (83%). Although 7 recently evaluated individuals received treatment based on the diagnosis provided and 5 individuals reportedly had "favorable" outcomes, a treatment benefit was not assessed.

Hainsworth and Greco (2016) retrospectively reviewed individuals with CUP who had the CancerTYPE ID test performed on tumor biopsies to identify those individuals predicted to have non-small cell lung cancer (NSCLC). GEP testing results showed that NSCLC was predicted in 37 of 310 individuals with CUP. A total of 21 of 37 individuals were tested for anaplastic lymphoma kinase (ALK) rearrangements, and the EML4-ALK fusion gene was detected in 4 individuals. A diagnosis of lung cancer was strongly suggested in only 1 individual prior to molecular testing. One individual received ALK inhibitor treatment with a reported “prolonged benefit.” The authors acknowledged that ALK inhibitors treatment experience and clinical data regarding the response of individuals with CUP to therapy targeted at molecular abnormalities identified by screening is limited.

RosettaGX Cancer Origin™ Test

Micro-RNAs (miRNAs) are small RNAs that act as regulators of messenger RNA (mRNA) to modulate the phases of cellular development, proliferation, and differentiation. MiRNAs have been suggested as potential biomarkers in identifying the tissue type from a given sample, and have been studied as a means to identify the tumor type and origin in individuals presenting with CUP when no primary site is apparent (Rosenfeld, 2008; Rosenwald, 2010). The miRview mets and miRview mets2) (Rosetta Genomics™ Laboratories, Philadelphia, PA; Rosetta Genomics Ltd., Rehovot, Israel) were developed as miRNA array-based assays used to identify 42 primary tumor types compared to a reference database of 336 tumors (range by tumor type, 1-49), with results reported as the tumor of origin or multiple possibilities. In an industry-sponsored validation study (Rosenwald, 2010), at least one of the two algorithms measuring expression levels of 48 miRNAs by qRT-PCR correctly predicted the reference diagnosis for the TOO in 85% (159 of 188) of the cases. In 66% of the cases (124 samples), a two algorithm prediction agreed on a single-tissue origin, which was identical to the reference diagnosis in 90% of cases. The authors stated the assay could provide a “necessary complement to diagnostic tools for the substantial number of patients that present with metastatic tumors, and thus save valuable time and facilitate treatment decisions. Further studies are underway to establish the ultimate clinical contribution and value of this test.” A second generation test, the RosettaGX Cancer Origin Test (formerly miRview mets2), expanded the number of tumor types to 42 primary origins with a panel of 64 miRNAs.

Varadhachary and colleagues (2011) prospectively studied the performance of the miRNA-based assay to identify the origin of metastatic disease in individuals with CUP. The TOO diagnosed was consistent with clinicopathologic features of the disease in 84% (62 of 74) of cases processed successfully (71% of all samples attempted). A total of 65 cases had differential diagnoses based on pathology and IHC results. The TOO assay was consistent with one of the differential diagnosis in 55 of these cases. Nine cases could not be classified by IHC or pathology. The miReview TOO results matched the clinical presentation for 7 of these cases. A limitation in the performance of the study includes the 16% disagreement of the level I or II results with the clinicopathologic findings. Because the assay may report two different diagnoses, a clinician may be presented with two different TOO predictions, including some in which the therapeutic management may differ significantly. In addition, tissue samples with origins that are not in the tumor panel cannot be classified correctly, as the assay can only report TOO that it has been trained to recognize. The investigators suggested that:

…comparative effectiveness research trials evaluating the added benefit of molecular profiling in appropriate CUP subsets are warranted. MicroRNA profiling may be particularly helpful in patients in whom the immunohistochemistry (IHC) profile of the metastasis is nondiagnostic or leaves a large differential diagnosis.

Mueller and colleagues (2011) attempted to validate the microRNA-based qRT-PCR test for identifying the TOO, first in a consecutive cohort of metastatic tumors of known origin and then in a cohort of CUP. Specimens were taken from 102 resected central nervous system (CNS) metastatic tumors with known origin, previously classified based on the person’s clinical history and pathological data, as well as a second cohort of resected CNS tumors from 57 persons originally diagnosed as CUP. For known tumors of the brain, 9 of the 12 cases of prostate cancer metastases (75%) were classified incorrectly and were subsequently analyzed separately. The reference diagnosis for TOO was predicted by at least one of the classifiers in 84% (75 of 89) of the remaining samples of known origin. For 52 of the 89 samples (58%), the two classifiers agreed, generating a consensus TOO prediction. Despite these lower prediction rates, the investigators suggested that the vast majority of samples for which the test was developed were primaries and metastases not from the CNS, and the pattern and overall performance values were similar to the results in the validation study (Rosenwald, 2010). In the second CUP cohort (57 brain and spinal metastatic samples from 54 different individuals), two classifiers were in agreement with 27 of the 54 samples (50%). It was noted that 3 cases were removed from the results in order to avoid duplication in reporting. To evaluate the performance of the test, the investigators developed a concordance score based on the clinicopathological data available at the time of diagnosis and additional information gathered after the test result was obtained. For four samples, no suggested origin existed and an “unknown” score was given. The test result predicted a suggested origin in 40 of the 50 samples (80%) that had a suggested origin based on clinical and/or pathological data. Limitations of this study include the exclusion of the prostate cancer metastases samples from further analyses because the investigators found that the correlation between the primary and metastatic prostate tissue was significantly lower than seen in other tissues. The investigators accounted for this exclusion by stating that prostate cancer is not considered a common TOO and represents only 2% of CUP cases. Additional study is needed to demonstrate the accuracy and clinical application of the microRNA test in identifying the TOO of metastases of unknown primary.

ProOnc TumorSourceDX™ Test

The ProOnc TumorSourceDX (Prometheus Laboratories Inc., San Diego, CA) uses technology similar to the miRview mets technology to identify TOO for a metastatic tumor. Tumor samples that are utilized are FFPE tumor sections from CUP sites. RNA is extracted from FFPE blocks using a proprietary method, which preserves miRNA. The RNA is then hybridized to miRNA probes adhered to a microarray platform. The test identifies 25 different tumor types, including colon, liver, brain, breast, kidney, lung, ovary, pancreas, prostate and testis, and measures the expression level of 48 miRNA biomarkers based on analysis of the microarray data using a decision tree classification algorithm. There are no studies currently available in the peer-reviewed medical literature to determine the safety and efficacy of the ProOnc TumorSourceDX technology to assist in clinical decision making.

Other Considerations

The Agency for Healthcare Research and Quality (AHRQ) (Meleth, 2013) published a technology assessment on the commercially available genetic tests used to identify the TOO in individuals with CUP site. The assessment included systematic reviews, randomized and nonrandomized controlled trials, prospective and retrospective cohort studies, case-control studies, and case series that assessed the analytical and clinical validity and clinical utility of the CancerTYPE ID, miReview, and Pathwork TOO tests. The assessment found that the accuracy rate across all studies for each of the three tests was fairly consistent, ranging from 84% for CancerTYPE ID assay, 87% for miRview mets, and 87% for Pathwork TOO; however, “the accuracy of the tests in CUP cases is not easily determined, because actual TOO is not identified in most cases.” The evidence that the TOO tests contribute to the diagnosis of CUP was moderate, with insufficient evidence to assess the effect of the tests on making clinical treatment decisions or outcomes, or to identify a survival benefit. Because the literature on molecular genetic tests for CUP is in its infancy, and all but one of the manuscripts reviewed were funded wholly or partially by the manufacturers of the tests, the assessment concluded that “the most urgent need in the literature is to have the clinical utility of the tests evaluated by research groups that have no evident conflict of interest.”

In 2015, the European Society of Medical Oncology (Fizazi, 2015) published a clinical practice guideline for the diagnosis, treatment and follow-up of CUP site. For GEP assays to identify TOO in individuals with CUP, the guideline states:

…their impact on patient outcome via administration of primary site specific therapy remains questionable and unproven in randomized trials (Level of evidence: IV based on retrospective cohort studies or case–control studies; Grade of recommendation C: insufficient evidence for efficacy or benefit does not outweigh the risk or the disadvantages)…Immunohistochemistry should be applied meticulously in order to identify the tissue of origin and to exclude chemosensitive and potentially curable tumors (ie, lymphomas and germ cell tumors).


The peer-reviewed medical literature contains studies addressing the reliability and validity of the TOO Test, CancerTYPE ID assay, and RosettaGX Cancer Origin Test; however, available data is limited to a small number of studies with variation in the tissue reference panels. Tissue reference standards are incomplete and the actual TOO may not be part of the cell types included in the reference panel. TOO analysis accuracy is further challenged when assessing poorly differentiated, metastatic tissues and particular tissue types. There are no published studies with a sufficient sample size that compare outcomes between individuals who received tissue-specific therapy after GEP identified TOO and those who have not. Further study is needed in the form of prospective trials to determine how GEP testing will impact clinical decision-making and outcomes of individuals with CUP.


CUP or occult primary malignancies are tumors that are manifested by a wide variety of clinical presentations and have a poor prognosis in most individuals (NCCN, V1.2019). The exact incidence of CUP in the United States is not precisely known. It is almost certainly underreported, and its true incidence is most probably between 2% and 6% of all diagnosed cancers (Siegel, 2012).

Most CUP can be classified into four major subtypes, the most frequently occurring is well or moderately differentiated adenocarcinoma (60%), followed by poorly differentiated adenocarcinoma or undifferentiated carcinoma (30%), squamous cell carcinoma (5%), and poorly differentiated malignant neoplasm (5%). Additionally, because of improved histopathologic diagnostic studies, neuroendocrine tumors of CUP have been recognized (1%) (NCCN, V1.2019). The most common primary sites of CUP are lung and pancreas, followed by colon and stomach, then breast, ovary, prostate, and solid-organ carcinomas of the kidney, thyroid, and liver. Metastatic disease is observed most commonly in the liver, lungs, bones, and lymph nodes. Conventional methods used to aid in the identification of the origin of a CUP include a thorough history and physical examination, computed tomography (CT) scans of the chest, abdomen, and pelvis; routine laboratory studies (including IHC staining); and targeted evaluation of specific signs and symptoms (Oien, 2008). Positron emission tomography (PET) may be used in the diagnosis when the tumor in question is a suspected or proven malignancy from a CUP. The current success rate of the diagnostic workup of a CUP is 20% to 30%, including consideration of clinical, radiologic, and extensive histopathologic methods (Horlings, 2008).

MicroRNAs (miRNAs) are small molecules of RNA, 21 to 23 nucleotides in length, that do not code for proteins but function to regulate expression of essential genes that play a role in cell development, differentiation, proliferation, and apoptosis. The GEP or signatures of miRNAs in tumors can be used to determine the original tissue type of CUP sites. miRNA expression profiling is most often performed using microarray assays, while alternative methods often include less high-throughput methods, such as the quantitative RT-PCR, or bead-based flow cytometry. RNA is typically extracted from samples of FFPE tumor sections of CUP. The origin of tumor is then determined using microarray-based miRNA expression profiling and a decision tree classification algorithm. The testing is usually considered for individuals with CUP to diagnosis primary origin or in those with cancers of suspected or known primary origin to confirm diagnosis.

Description of the GEP Tests

In July 2008, the Pathwork TOO test (now known as the TOO Test) was cleared by the FDA through the 510(k) process for use in measuring:

the degree of similarity between the RNA expression pattern in a patient’s fresh-frozen tumor and the RNA expression patterns in a database tumor samples (poorly differentiated, undifferentiated and metastatic cases) that were diagnosed according to the current clinical and pathological practice.

The test measures the expression of more than 1500 genes and compares the similarity of the GEP of a CUP to a database of known profiles from 15 tissues with more than 60 histologic morphologies. The report generated for each tumor consists of a “similarity score,” which is a measure of similarity of the GEP of the specimen to the profile of the 15 known tumors in the database. Scores range from 0 (very low similarity) to 100 (very high similarity), and sum to 100 across all 15 tissues on the panel. If a single similarity score is ≥ 30, it indicates that this is likely the TOO. If every similarity score is between 5 and 30, the test result is considered indeterminate, and a similarity score of < 5 rules out that tissue type as the likely origin.

The test was the first of its kind to receive FDA clearance for use in analyzing a tumor’s gene expression pattern to help pinpoint the source of hard-to-identify tumors. The test result “is intended for use in the context of the patient's clinical history and other diagnostic tests evaluated by a qualified clinician.” In June 2010, the FDA expanded clearance for the test to be run on an individual’s formalin fixed, paraffin-embedded (FFPE) tumor. Limitations to the FDA clearance for both tests are as follows:

The Pathwork is not intended to establish the origin of tumors (e.g. cancer of unknown primary) that cannot be diagnosed according to current clinical and pathological practice. It is not intended to subclassify or modify the classification of tumors that can be diagnosed by current clinical and pathological practice, or to predict disease course or survival or treatment efficacy, or to distinguish primary from metastatic tumor. Tumor types not in the Pathwork Tissue of Origin Test database may have RNA expression patterns that are similar to patterns in the database; therefore, results cannot be used to distinguish tumor types in the database from tumor types not in the database (FDA, 2010).

On March 15, 2018, the FDA cleared modifications to the TOO Test Kit-FFPE (Cancer Genetics, Inc., Rutherford, NJ) (K173839) with the same indications and intended use as the predicate device, the Pathwork TOO FFPE (K092967 and K120489). Device changes include use of a different in vitro diagnostic product plus microchip and software modifications.

The CancerTYPE ID RT-PCR assay is a GEP test that provides molecular classification of cancers with indeterminate, uncertain, or differential diagnoses. The test uses a proprietary algorithm to compute a probability for each tumor type, and the tumor type with the highest probability is predicted to be the most likely source for the tumor. According to the manufacturer:

CancerTYPE ID is not intended to predict patient survival benefit, treatment efficacy or to distinguish between benign versus malignant lesions. Tumor types not included in the CancerTYPE ID reference database may exhibit RNA expression patterns that are similar to RNA expression patterns within the reference database.

The CancerTYPE ID test is performed at bioTheranostics’ CLIA-certified, clinical laboratory in San Diego, California. The RT-PCR can be used at the practice level; however, it can only measure, at most, a few hundred genes, limiting tumor categorization to seven or fewer types.

To date, neither the CancerType ID assay nor the RosettaGX Cancer Origin Test (miRview mets2 assay) have been submitted to the FDA for 510(k) clearance.


Cancer of unknown primary (CUP): A case in which cancer cells are found in the body, but the place where the cells first started growing (the origin or primary site) cannot be determined. Also called carcinoma of unknown primary and CUP.

Cytology: The study of the formation and function of cells.

Deoxyribonucleic acid (DNA): The hereditary material in humans and almost all other organisms. Genetic material contained in nearly every cell in a person’s body has the same DNA.

Gene expression profile/profiling (GEP): Information about all messenger RNAs that are made in various cell types. A gene expression profile may be used to find and diagnose a disease or condition and to see how well the body responds to treatment.

Histology: The study of the microscopic structure of tissue and cells.

Microarray-based gene expression profile: The ability to measure and analyze thousands of genes simultaneously in a single RNA sample; also referred to as gene expression microarray (GEM).

Ribonucleic acid (RNA): A nucleic acid found that transmits genetic instructions from the nucleus to the cytoplasm of cells; functions in the assembly of proteins.

Sensitivity of a clinical test: Refers to the ability of the test to correctly identify those individuals with the disease.

Specificity of a clinical test: Refers to the ability of the test to correctly identify those individuals without the disease.



The following codes for treatments and procedures applicable to this document are included below for informational purposes. Inclusion or exclusion of a procedure, diagnosis or device code(s) does not constitute or imply member coverage or provider reimbursement policy. Please refer to the member's contract benefits in effect at the time of service to determine coverage or non-coverage of these services as it applies to an individual member.

When services are Investigational and Not Medically Necessary:




Oncology (tissue of origin), microarray gene expression profiling of > 2000 genes, utilizing formalin-fixed paraffin-embedded tissue, algorithm reported as tissue similarity scores
Pathwork Tissue of Origin Test, Pathwork Diagnostics


Oncology (tumor of unknown origin), mRNA, gene expression profiling by real-time RT-PCR of 92 genes (87 content and 5 housekeeping) to classify tumor into main cancer type and subtype, utilizing formalin-fixed paraffin-embedded tissue, algorithm reported as a probability of a predicted main cancer type and subtype
CancerTYPE ID, bioTheranostics, Inc.


Unlisted multianalyte assay with algorithmic analysis [when specified as a test involving gene expression profiling for cancers of unknown primary]



ICD-10 Diagnosis



All diagnoses

When services are also Investigational and Not Medically Necessary:
When the code describes a procedure indicated in the Position Statement section as investigational and not medically necessary.




Molecular pathology procedure, Level 7 (eg, analysis of 11-25 exons by DNA sequence analysis, mutation scanning or duplication/deletion variants of 26-50 exons, cytogenomic array analysis for neoplasia) [when specified as the following]:

  • Cytogenomic microarray analysis, neoplasia (eg, interrogation of copy number, and loss-of-heterozygosity via single nucleotide polymorphism [SNP]-based comparative genomic hybridization [CGH] microarray analysis)



ICD-10 Diagnosis



For the following diagnoses to distinguish primary versus metastatic or to establish site or origin:


Malignant neoplasms


Malignant (primary) neoplasm, unspecified


Malignant neoplasms of lymphoid, hematopoietic and related tissue


Peer Reviewed Publications:

  1. Anderson GG, Weiss LM. Determining tissue of origin for metastatic cancers: meta-analysis and literature review of immunohistochemistry performance. Appl Immunohistochem Mol Morphol. 2010; 18(1):3-8.
  2. Beck AH, Rodriguez-Paris J, Zehnder J, Schrijver I. Evaluation of a gene expression microarray-based assay to determine tissue type of origin on a diverse set of 49 malignancies. Am J Surg Pathol. 2011; 35(7):1030-1037.
  3. Bender RA, Erlander MG. Molecular classification of unknown primary cancer. Semin Oncol. 2009; 36(1):38-43.
  4. Blaszyk H, Hartmann A, Bjornsson J. Cancer of unknown primary: clinicopathologic correlations. APMIS. 2003; 111(12):1089-1094.
  5. Brachtel EF, Operana TN, Sullivan PS, et al. Molecular classification of cancer with the 92-gene assay in cytology and limited tissue samples. Oncotarget. 2016; 7(19):27220-27231.
  6. Dumur CI, Fuller CE, Blevins TL, et al. Clinical verification of the performance of the pathwork tissue of origin test: utility and limitations. Am J Clin Pathol. 2011; 136(6):924-933.
  7. Dumur CI, Lyons-Weiler M, Sciulli C, et al. Interlaboratory performance of a microarray-based gene expression test to determine tissue of origin in poorly differentiated and undifferentiated cancers. J Mol Diagn. 2008; 10(1):67-77.
  8. Erlander MG, Ma XJ, Kesty NC, et al. Performance and clinical evaluation of the 92-gene real-time PCR assay for tumor classification. J Mol Diagn. 2011; 13(5):493-503.
  9. Greco FA, Erlander MG. Molecular classification of cancers of unknown primary site. Mol Diagn Ther. 2009; 13(6):367-373.
  10. Greco FA, Lennington WJ, Spigel DR, Hainsworth JD. Molecular profiling diagnosis in unknown primary cancer: accuracy and ability to complement standard pathology. J Natl Cancer Inst. 2013; 105(11):782-790.
  11. Greco FA, Lennington WJ, Spigel DR, Hainsworth JD. Poorly differentiated neoplasms of unknown primary site: diagnostic usefulness of a molecular cancer classifier assay. Mol Diagn Ther. 2015; 19(2):91-97.
  12. Grenert JP, Smith A, Ruan W, et al. Gene expression profiling from formalin-fixed, paraffin-embedded tissue for tumor diagnosis. Clin Chim Acta. 2011; 412(15-16):1462-1464.
  13. Hainsworth JD, Greco FA. Lung adenocarcinoma with anaplastic lymphoma kinase (ALK) rearrangement presenting as carcinoma of unknown primary site: recognition and treatment implications. Drugs Real World Outcomes. 2016; 3(1):115-120.
  14. Hainsworth JD, Rubin MS, Spigel DR, et al. Molecular gene expression profiling to predict the tissue of origin and direct site-specific therapy in patients with carcinoma of unknown primary site: a prospective trial of the Sarah Cannon Research Institute. J Clin Oncol. 2013; 31(2):217-223.
  15. Hainsworth JD, Schnabel CA, Erlander MG, et al. A retrospective study of treatment outcomes in patients with carcinoma of unknown primary site and a colorectal cancer molecular profile. Clin Colorectal Cancer. 2012; 11(2):112-118.
  16. Handorf CR, Kulkarni A, Grenert JP, et al. A multicenter study directly comparing the diagnostic accuracy of gene expression profiling and immunohistochemistry for primary site identification in metastatic tumors. Am J Surg Pathol. 2013; 37(7):1067-1075.
  17. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer J Clin. 2010; 60(5):277-300.
  18. Kerr SE, Schnabel CA, Sullivan PS, et al. Multisite validation study to determine performance characteristics of a 92-gene molecular cancer classifier. Clin Cancer Res. 2012; 18(14):3952-3960.
  19. Monzon FA, Koen TJ. Diagnosis of metastatic neoplasms: molecular approaches for identification of tissue of origin. Arch Pathol Lab Med. 2010; 134(2):216-224.
  20. Monzon FA, Lyons-Weiler M, Buturovic LJ, et al. Multicenter validation of a 1,550-gene expression profile for identification of tumor tissue origin. J Clin Oncol. 2009; 27(15):2503-2508.
  21. Mueller WC, Spector Y, Edmonston TB, et al. Accurate classification of metastatic brain tumors using a novel microRNA-based test. Oncologist. 2011; 16(2):165-174.
  22. Oien KA, Evans TR. Raising the profile of cancer of unknown primary. J Clin Oncol. 2008; 26(27):4373-4375.
  23. Pillai R, Deeter R, Rigl CT, et al. Validation and reproducibility of a microarray-based gene expression test for tumor identification in formalin-fixed, paraffin-embedded specimens. J Mol Diagn. 2011; 13(1):48-56.
  24. Rosenfeld N, Aharonov R, Meiri E, et al. MicroRNAs accurately identify cancer tissue origin. Nat Biotechnol. 2008; 26(4):462-469.
  25. Rosenwald S, Gilad S, Benjamin S, et al. Validation of a microRNA-based qRT-PCR test for accurate identification of tumor tissue origin. Mod Pathol. 2010; 23(6):814-823.
  26. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin. 2012; 62(1):10-29.
  27. Stancel GA, Coffey D, Alvarez K, et al. Identification of tissue of origin in body fluid specimens using a gene expression microarray assay. Cancer Cytopathol. 2012; 120(1):62-70.
  28. Tothill RW, Kowalczyk A, Rischin D, et al. An expression-based site of origin diagnostic method designed for clinical application to cancer of unknown origin. Cancer Res 2005; 65(10):4031-4040.
  29. Varadhachary GR, Spector Y, Abbruzzese JL, et al. Prospective gene signature study using microRNA to identify the tissue of origin in patients with carcinoma of unknown primary. Clin Cancer Res. 2011; 17(12):4063-4070.

Government Agency, Medical Society, and Other Authoritative Publications:

  1. Fizazi K, Greco FA, Pavlidis N, et al; ESMO Guidelines Committee. Cancers of unknown primary site: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2015; 26 Suppl 5:v133-v138.
  2. Meleth S, Whitehead N, Swinson T, Lux L. Technology assessment on genetic testing or molecular pathology testing of cancers with unknown primary site to determine origin. Technology Assessment Report. Final. Prepared by RTI International-University of North Carolina at Chapel Hill Evidence-based Practice Center for the Agency for Healthcare Research and Quality (AHRQ) under contract no. HHSA290200710057I #8. Project ID: CANU5011. Rockville, MD: AHRQ; February 20, 2013. Available at: Accessed on September 13, 2018.
  3. NCCN Clinical Practice Guidelines in Oncology®. ©2018 National Comprehensive Cancer Network®, Inc. Occult Cancer (Cancer of Unknown Primary [CUP]). V1.2019. Revised August 3, 2018. For additional information visit the NCCN website: Accessed on September 13, 2018.
  4. U.S. Food and Drug Administration (FDA). 510(k) Premarket Notification Database. Product Code: OIW. Available at: Accessed on September 13, 2018.
Websites for Additional Information
  1. National Cancer Institute (NCI). Cancer of Unknown Primary Treatment (PDQ®). Last modified February 4, 2018. Available at: Accessed on September 13, 2018.

Formalin-Fixed Paraffin-Embedded (FFPE)
ProOnc TumorSourceDX
Real-Time Polymerase Chain Reaction (RT-PCR)
RosettaGX Cancer Origin
Tissue of Origin Test
Tissue of Origin Test-FFPE

The use of specific product names is illustrative only.  It is not intended to be a recommendation of one product over another, and is not intended to represent a complete listing of all products available. 

Document History






Medical Policy & Technology Assessment Committee (MPTAC) review.



Hematology/Oncology Subcommittee review. Updated Rationale, Background, References, Websites for Additional Information, and Index sections.



MPTAC review.



Hematology/Oncology Subcommittee review. The document header wording updated from “Current Effective Date” to “Publish Date.” Updated Rationale, Background, References, and Websites for Additional Information sections.



MPTAC review.



Hematology/Oncology Subcommittee review. Updated formatting in Position Statement section. Updated Rationale, Background, References, Websites for Additional Information, and Index sections.



MPTAC review.



Hematology/Oncology Subcommittee review. Updated Rationale and References section.  Updated Coding section with 01/01/2016 CPT changes; also removed ICD-9 codes.



MPTAC review.



Hematology/Oncology Subcommittee review. Updated Rationale, Background, References, Websites for Additional Information, and Index sections.



MPTAC review.



Hematology/Oncology Subcommittee review. Updated Rationale, References, and Websites for Additional Information sections. Updated Coding section with 01/01/2014 CPT changes.



MPTAC review.



Hematology/Oncology Subcommittee review. Clarified Position Statement. Updated Description, Rationale, Discussion, Definitions, References, Websites for Additional Information and Index.  Updated Coding section with 01/01/2013 CPT changes; removed 88384-88386 deleted 12/31/2012.



MPTAC review.



Hematology/Oncology Subcommittee review. Revised the Subject of the document. Revised investigational and not medically necessary Position Statement to address GEP testing techniques for evaluating CUP. Updated Description, Rationale, Background, Definitions, References, Websites for Additional Information, and Index.  Updated Coding section including 01/01/2012 CPT changes.



MPTAC review. Initial document development.