Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards

ASCO University
Aug 12, 2015 9:39 AM

User Instructions: Welcome to the Molecular Oncology Tumor Board Series! This educational initiative is a collaboration of the American Society of Clinical Oncology (ASCO), College of American Pathologists (CAP), and Association for Molecular Pathology (AMP).

This month’s topic on carcinoma of unknown primary is led by Drs. F. Anthony Greco (Sarah Cannon Research Institute and Tennessee Oncology) and Valentina Nardi (Massachusetts General Hospital).

New: There will be two related patient cases posted this month! The second case will be posted August 17th at the time of faculty responses to Case #1.

Users are encouraged to leave comments and post questions about the case in order to generate a wide discussion among the cancer care community. You can also receive email notifications when new comments are posted by clicking the “Follow this Conversation” option located at the bottom of this page.

Supplemental resources related to the case are attached for reference.

Comments

11081

ASCO University
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 12, 2015 9:43 AM

Patient Case #1

Medical History and Initial Findings 

A 26 year old woman (never smoker) had RLQ abdominal pain which was persistent and severe. CT scans of the chest, abdomen/pelvis showed an ovarian mass and PET scan revealed increased uptake in the mass as well as in normal-sized right supraclavicular and mediastinal nodes and the right sixth rib.  Physical examination was unremarkable as were mammograms, esophagogastroduodenoscopy and colonoscopy.  She underwent resection of the mass with right unilateral salpingo-oophorectomy.

Laboratory Studies

The ovarian mass was an adenocarcinoma with some areas of mucin production. This was felt to be a metastatic lesion to the ovary (Krukenberg tumor). Immunohistochemical staining revealed CK-7 positive and CK-20, TTF-1, ER/PR, WT-1, CD-56 and chromogranin were all negative. Serum CEA was 40 and CA-125 normal

Diagnosis 

She was diagnosed with a metastatic adenocarcinoma of unknown primary site involving her right ovary and right lateral sixth rib with suspicion of metastatic disease in the supraclavicular and mediastinal nodes.

Additional imaging related to the case can be accessed here

Clinical Course

It was felt that she may have metastasis from an occult primary adenocarcinoma and she was treated initially with paclitaxel and carboplatin for several cycles and radiation therapy to the right sixth rib. She had evidence of progression in retroperitoneal lymph nodes after 19 months. Needle biopsy of the nodal mass showed adenocarcinoma, similar to previous pathology. She was treated with modified FOLFOX chemotherapy and responded well. The biopsy tissue block from the retroperitoneal node biopsy was sent for the RT PCR 92 gene molecular cancer classifier assay (CancerTYPE ID®) and returned as a 90% probability of adenocarcinoma of the lung. EGFR was found to be wild type and ALK by FISH revealed the classic rearrangement EML 4-ALK fusion gene.  Modified FOLFOX was continued for several cycles with a complete response. Upon progression she will be considered a candidate for an ALK inhibitor. 

Additional imaging related to the case can be accessed here.

11086

ASCO University
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 12, 2015 9:44 AM

Discussion Questions

  1. What tumor types should be considered in the differential diagnosis of a CK7+ CK20- CUP?
  2. Could you have performed additional immunohistochemical stains to try to determine the site of origin?
  3. Is EML4-ALK fusion unique to lung adenocarcinomas?
  4. Does genetic testing (molecular cancer classifier assay and comprehensive genetic profiling) play a role in accurately identifying the cancer type/site of origin and guiding therapy for CUP patients?

 

11091

Anis Toumeh, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 12, 2015 6:41 PM

This is an interesting case, indeed.

1- With the CK7+ CK20- CUP, one should conisder lung, breast, upper GI, pancreatobiliary and mullerian as possible primary sites.

2- Additional testing of PAX8, CDX2 and Napsin-A may have helped.

3- The ALK gene was originally identified in a subset of anaplastic large cell lymphomas with a t( 2;5)( p23; q35) translocation. In  NSCLC, chromosome 2p inversion results in fusion of the protein encoded by the echinoderm microtubule-associated protein-like 4 (EML4) gene with the intracellular signaling portion of the ALK receptor tyrosine kinase.

4- I believe genetic testing does play a role when trying to identify the tissue of origin or, more importantly, whether or not there is an actionable genetic alteration like in the patients presented. 

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 12, 2015 10:21 PM

Course Faculty Response

Thank you for your informative comments Dr. Toumeh. Immunohistochemistry is very important in helping to diagnose the cancer type in CUP and the additional stains you mentioned can be helpful when considering ovarian (PAX8), gastrointestinal cancers (CDX2) and lung (Napsin-A). Dr. Nardi may want to expand on this issue. Molecular profiling with a molecular cancer classifier assay is now clearly useful in diagnosis of the cancer type and complements immunochemistry. CUP cancers appear to  respond to therapies predictably according to their origin.  If a carcinoma of the lung is the likely diagnosis based on immunohistochemistry and/or a molecular cancer classifier assay one would want to check for EGFR mutations and ALK/ROS1 rearrangements as was done in these patients. We may one day use comprehensive molecular profiling in all advanced cancer patients regardless of their cancer type to determine therapy. We have not yet reached that day for most patients.

 

Valentina Nardi
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 13, 2015 9:41 AM

Course Faculty Response

Thank you Dr. Toumeh for your comments. Indeed a more extensive panel of immunohistochemical stains, to include additional tissue-restricted markers (eg. Napsin A for lung, PAX8 for ovary, GATA-3 for breast and urothelial, CDX-2 for gastrointestinal adenocarcinomas) could be performed in similar cases, when the initial panel of immunohistochemical stains does not point to a specific tissue of origin.

When the sample is limited (small biopsy or cell block), and in the absence of clinical or radiological evidence of a primary lesion, pathologists often try to preserve tissue for molecular studies, which can include gene expression profile to identify the tissue of origin and/or tumor genotyping  with detection also of translocations  to identify actionable genetic alterations.

 

11106

Shayan Kaka Salh Al-Askary, MBChB
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 13, 2015 10:15 AM

She has midline adenocarcinoma and young age 26 years old What about tumor markers for extragonadal germ cell tumor like AFP & B-HCG

Valentina Nardi
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 13, 2015 10:41 AM

Course Faculty Response

Thank you for your comment Dr. Shayan Kaka Salh Al-Askary.

While based on the clinical presentation, a germ cell tumor may be a consideration (and I will let Dr. Greco further comment on this), the histological appearance of the tumor was not consistent with that of a germ cell tumor.

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 13, 2015 12:51 PM

Course Faculty Response

I agree with Dr. Nardi. A mucin positive adenocarcinoma even as a component of a germ cell tumor would be exceedingly unlikely. However, since she presented with a large ovarian mass at a young age it would be have been reasonable to check the tumor markers before her surgery.

11121

Gary Spitzer, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 13, 2015 6:41 PM

it is still confusing t me given the extra tissue at rebiopsy that the likely orgin could not have been determined by extra immunochemistry, ca 19-9 levels, and then given non smoker status reflex to egfr and alk and ros testing.

what features on this cancer idx test are so are so diagnostic of an adenocarcinoma TTF1 negative cd7+ from an upper GI tumor

Valentina Nardi
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 14, 2015 9:04 AM

Course Faculty Response

Dr. Spitzer, thank you for your comment. In retrospect, it is possible that additional immunohistochemistry and negative tumor markers might have helped  narrowing down the diagnosis.

The clinical scenario was not very suggestive of lung cancer, which, very rarely has been reported to metastatize to ovaries in the absence of an identifiable lung mass (Irving JA and Young RH, AM J Surg Path 2005 Aug;29(8):997-1006). I am not sure I am interpreting correctly your second statement, but I agree that an adenocarcinoma TTF1- CK7+ CK20- could certainly be of upper GI origin.

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 14, 2015 9:41 AM

Course Faculty Response

The immunohistochemical (IHC) evaluation was not highly suggestive of a single cancer type. IHC profiles or patterns are useful in diagnosis for many CUP cancers but in at least 50% a single diagnosis is not possible. The positive CK7 stain is nonspecific but is usually positive in upper GI, breast, lung, ovary and others. The TTF-1 stain is positive in 80% of lung adenocarcinomas but was negative in both her biopsy specimens. Her anatomical primary site was not clinically detected ( thus the diagnosis of CUP) and standard pathology on the metastatic lesions was unable to define a likely single type of adenocarcinoma. If the TTF-1 had been unequivocally positive the diagnosis of lung adenocarcinoma would have been very likely and the molecular cancer classifier assay would not have been necessary. The 92-gene RT-PCR molecular cancer classifier was designed to determine or diagnose the specific cancer type based upon patterns of gene expression specific to the tissue of origin and is about 90% accurate. This assay frequently can diagnose the cancer type when IHC is inconclusive. The 92-gene RT-PCR performed on her biopsy showed a 90% probability of lung adenocarcinoma.

She very likely has an occult (very small clinically undetectable) lung adenocarcinoma. A focused molecular evaluation is now standard for lung adenocarcinoma looking for EGFR mutations and ALK/ROS1 rearrangements since useful targeted therapies are available for patients with these genetic alterations. The positive FISH test for ALK rearrangement on her biopsy further supports the diagnosis of lung adenocarcinoma. Although her clinical presentation was atypical for lung cancer, which is common in CUP, she has some features associated with ALK rearranged lung adenocarcinoma including her age, never smoker status and mucin producing adenocarcinoma. Most CUP patients have an occult primary cancer and the use of IHC and molecular diagnostic studies are now capable of determining the specific cancer type of most tumors by testing a biopsy from a metastatic site.

11136

MOHAMED YOUSSEF DEIBAS, MBBS
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 14, 2015 11:03 PM

thank you for your informative participation but if we see the patern of the disease we found that 1st posibility of ovarian cancer, 2nd posibility is colorectal carcinoma also can spread through peritoneium and lymphatics. lung adenocarcinoma different completely also respons on chemotherapy away from lung, if the patient showing response on folfox can we asses kras status

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 15, 2015 3:14 PM

Course Faculty Response

I agree that her clinical presentation or pattern of the disease would first suggest primary ovarian cancer and if not, metastatic carcinoma from the gastrointestinal tract (colorectal or gastric most likely). The metastatic patterns in the minority of CUP patients are atypical from that expected from their counterparts with detectable primary sites. A known lung cancer may metastasize to the ovary infrequently, but when the primary is very small and clinically undetectable the pattern of metastasis seems to inexplicably differ at least for some CUP patients. Occult prostate adenocarcinoma seems to spread to nodes and lung as often as bone; occult pancreatic adenocarcinoma spreads to bone more frequently than overt primaries. I suspect there is a genetic and/or epigenetic mechanism but the explanation remains a mystery as does the reason the primary remains occult yet produces clinically detectable metastasis.

It does also seem unusual that she has responded exceptionally well, particularly to a FOLFOX regimen but the cytotoxic effect is rather nonspecific and most data are based on large numbers of treated patients rather than single examples. If she relapses I would expect a good response to an ALK inhibitor.

Her adenocarcinoma did not have the IHC pattern of ovarian or gastrointestinal adenocarcinoma. The gene expression was that of adenocarcinoma of the lung and there was an ALK rearrangement. All the clinical, pathological and genetic data diagnose her with metastatic lung adenocarcinoma. She has the CUP syndrome but her cancer type has been determined. A KRAS could be checked but the clinical importance is uncertain.

 

11146

Mayer Gorbaty, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 15, 2015 2:21 PM

I have two management questions: 

1.  Once the Cancertype ID came back lung cancer and the EML4/ALK rearrangement was identified would you still have continued the FOLFOX as done here or would you have stopped it to avoid toxicty and switch immediately to crizotinib?

2.  Now that she is doing great and is in CR- do you agree with waiting for this otherwise healthy young woman to relapse before starting crizotinib or might you go ahead and start her before evidence of relapse?

 

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 15, 2015 11:33 PM

Course Faculty Response

I would have continued FOLFOX since she was responding and tolerating well. Neither FOLFOX or crizotinib is curative therapy in the setting. Resistance invariably eventually develops. She is off all therapy in CR now with a good quality of life. When she has tumor progression an ALK inhibitor would be my first choice. Whether her ultimate survival would be better by immediate use of an ALK inhibitor or awaiting relapse in not known. Giving her crizotinib now seems reasonable as does awaiting relapse.

11161

Maureen A. Cooper, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 17, 2015 2:34 PM

Why folfox when the data appears to demonstrate lung primary?

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 17, 2015 5:01 PM

Course Faculty Resposne

Her clinical course has occurred during the last several years during the evolving knowledge regarding diagnosis of the cancer type and treatment of CUP. The initial empiric treatment with taxol/carboplatin produced a long response. We felt later she most likely had a GI occult primary and FOLFOX was started before we had the molecular cancer classifier assay and ALK rearrangement results. She responded so well to FOLFOX we decided to not use lung directed therapy but are poised to do so when she relapses. In current practice if one has contemporary similar information the treatment should be site directed for lung adenocarcinoma.

11166

Valentina Nardi
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 17, 2015 3:14 PM

Course Faculty Response

  1. The following tumor types would enter in the differential diagnosis of a CK7+ CK20- carcinoma: Lung adenocarcinoma, breast, thyroid carcinoma, endometrial, cervical, salivary gland and, pancreatobiliary, gastric, esophageal. 

    The results of the additional immunohistochemical stains performed revealed no staining for CK20, TTF1, ER/PR, WT-1, CD56 and chromogranin, making it less likely to be a metastasis from a lung, thyroid, breast or ovarian primary. 

    Keratin 7 and keratin 20 are widely used to predict primary site (Chu P, Wu E, Weiss LM, Mod Pathol 2000; 13:962-972).

    Although the CK7+ CK20- immunophenotype is most common in CUP, it is not particularly helpful in suggesting a specific anatomic site of origin.
     

  2. Additional immunohistochemical stains (e.g. NapsinA for lung, GATA-3 for breast and urothelial, CDX-2 for gastrointestinal adenocarcinomas) were not performed at the time, in the absence of clinical or radiological evidence of a primary lesion, and trying to preserve tissue for molecular studies.
     
  3. EML4-ALK fusion is not unique to non-small cell lung cancer. This fusion has been detected, at lower frequency, in breast, colorectal adenocarcinomas (PMID 19737969), although its role in the pathogenesis of these tumors is unknown.
     
  4. Comprehensive genetic profiling, looking at single nucleotide variants, insertions and deletions, copy number changes and gene fusions can identify actionable genetic alterations, and in some cases, can suggest an anatomic site of origin ( PMID: 23547070; 24037760; JAMA Oncol. 2015;1(1):40-49. doi:10.1001/jamaoncol.2014.216.)

11171

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 17, 2015 3:24 PM

Course Faculty Response

  1.  No additional commentary
     
  2.  No additional commentary
     
  3.  In her clinical setting, which is always important to consider in CUP, lung adenocarcinoma is the most likely diagnosis, although EML4-ALK fusion is not unique to lung adenocarcinomas. She had rather atypical metastasis from lung carcinoma to her ovary, but CUP patients often have metastasis to sites not usually associated with their site of origin/primary site.  She also had some typical metastasis for a lung primary to bone, mediastinal and supraclavicular and retroperitoneal lymph nodes. (Greco FA and Hainsworth JD. Cancer of Unknown primary Site. In Cancer: Principles and Practice of Oncology. Edited by DeVita VT, Lawerence TS, Rosenberg SA. Philadelphia, Lippincott, Williams, and Wilkins; Tenth edition 2015; 1720-1737.)

    If the TTF-1 stain of her cancer cells was positive, there would have been no need to obtain the molecular cancer classifier assay since this immunophenotype (CK7+, CK20-, TTF-1+) is almost always lung adenocarcinoma (rarely, other adenocarcinomas are possible, including thyroid).   The ALK rearrangement gave additional support to the diagnosis of lung carcinoma. The TTF-1 stain is negative in 15-20% of lung adenocarcinomas. The molecular cancer classifier assay revealed the gene expression profile of lung adenocarcinoma (90% probability). She otherwise also fits the clinicopathologic picture of ALK-rearranged lung adenocarcinomas (younger patient, non-smoker and mucin-producing adenocarcinoma). (Shaw AT, Yeap BY, Mino-Kenudson M, et al Clinical features and outcome of patients with non-small cell lung cancer who harbor EML4-ALK. J Clin Oncol 2009; 27: 4247-4253)  An anatomical primary cancer in the lung/bronchus or elsewhere was not identified thus defining her within the CUP syndrome. Autopsies of CUP patients has identified the primary site in about 75% and usually the primaries are very small (clinically occult) and not detectable by routine imaging studies or endoscopies. The explanation for this unusual biologic behavior is unknown but likely will eventually be explained by unique genetic and/or epigenetic abnormalities. (Greco FA. Cancer of unknown primary site; still an entity, a biologic mystery and a metastatic model. Nat Rev Cancer. 2014; 14(1);3-4.)

    In summary, the sites of metastasis, the morphology of the biopsy specimen, the molecular cancer classifier diagnosis and the ALK rearrangement make it very likely that she has lung adenocarcinoma arising from an occult lung/bronchial primary site.
     

  4. Molecular cancer classifier assays are about 90% accurate in determining the cancer type, both in patients with known cancer types (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: 3592-3560.) as well as those who have CUP (Greco FA, Lennington WJ, Spigel DR, et al. Molecular profiling diagnosis in unknown primary cancer: accuracy and ability to complement standard pathology. J Natl Cancer Inst. 2013, 5;105(11): 782-790.). The gene expression profiles of CUP cancers are rarely indeterminate, overlap with more than one cancer type or are cancers which are not included in the library/panel of cancers recognizable by the assay. The diagnosis made by a molecular cancer classifier assay should always be considered in concert with the clinical features including patient’s gender, evaluation looking for a primary site, metastatic sites, and morphology of the cancer along with immunohistochemistry findings. The vast majority (95%) of CUP patients may have their cancer type now diagnosed. (Hainsworth JD, Greco FA. Gene expression profiling in patients with carcinoma of unknown primary site: from translational research to standard of care. Virchows Arch. 2014. 464: 393-402.)  Accurate identification of the type/origin of the cancer helps guide appropriate site-directed therapy as well as prompting additional specific testing for associated actionable genetic alterations such as ALK fusion, as illustrated by both cases presented.

    Comprehensive genetic profiling or next generation sequencing platforms are designed to detect genetic alterations in cancer cells rather than the type of cancer, but occasionally the results (genetic alteration found) can suggest the cancer type/site of origin, such as the finding of an ALK arrangement or EGFR mutation in a carcinoma would highly suggest lung adenocarcinoma. To help determine the cancer type in CUP patients, a molecular cancer classifier is a more specific test to obtain, particularly if a reasonable number of immunohistochemical stains does not highly suggest/diagnose a single cancer type. Treating CUP patients with site specific therapies based on their immunohistochemical or molecular diagnosis of their cancer type improves their outcome, particularly for those patients with more responsive/treatable advanced cancers. (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 2012; 31; 217-223.)   It is not feasible or appropriate at this time to obtain comprehensive genetic profiling (next generation sequencing) in every CUP patient or, for that matter, in every patient with advanced known cancer of any type. Additional studies of targeted or precision therapies in patients with the appropriate target as determined by comprehensive genetic profiling are underway and may be useful for patients with many advanced solid tumors, including CUP.

11176

ASCO University
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 17, 2015 3:28 PM

Patient Case #2

Medical History

A 52 year old male previous cigarette smoker developed persistent cough and chest pain. He was found to have a right paratracheal mass by CT scan of the chest and increased uptake in the mass by PET scan. There were no other abnormalities by CT scans of the chest, abdomen and pelvis or by PET scanning. Bronchoscopy was unremarkable and a transbronchial biopsy of the mediastinal mass showed a poorly differentiated carcinoma. An anatomical lung primary site was not identified. He underwent thoracotomy and mediastinal lymphadenectomy; the nodal mass was 4 cm in diameter, replaced by metastatic poorly differentiated carcinoma. Multiple other nodes were negative.

Laboratory Studies

Immunohistochemical staining of the cancer cells were CK-7 positive, CK-20, TTF-1 negative and scattered staining with CD-117, HCG and PLAP. The pathologic diagnosis was a carcinoma not otherwise specified. Serum HCG and alpha fetoprotein were negative. A second pathologic consultation was obtained and several other stains including human placental lactogen, calretinin, WT-1 and podoplanin were negative. Other than carcinoma, no additional diagnosis was rendered. FISH testing of the cancer cells for chromosome 12 abnormalities was negative and the RT-PCR 92 gene molecular cancer classifier (CancerTYPE ID®) reported a 77% probability of lung adenocarcinoma. EGFR was wild type and ALK by FISH showed the classic rearrangement EML 4-ALK fusion.

Diagnosis

Carcinoma of unknown primary site involving the mediastinum

11181

ASCO University
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 17, 2015 3:30 PM

Discussion Questions

  1. What anatomical sites of origin are suggested by this immunohistochemical pattern?
  2. Does the genetic testing (molecular cancer classifier assay) and the ALK rearrangement found by FISH testing make the diagnosis of carcinoma of the lung secure?

11186

Anis Toumeh, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 17, 2015 3:46 PM

That is rather an interesting IHC panel. What do you make of the "scattered" PLAP and HCG? Should we consider Germ Cell?

Valentina Nardi
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 18, 2015 9:03 AM

Course Faculty Response

Dr. Toumeh, indeed the positive staining for PLAP and HCG raises the possibility of a germ cell tumor. PLAP can be elevated in smokers and therefore in this setting the PLAP staining is non-specific and morphological and further serological and immunohistochemical studies showed no evidence of a germ cell tumor.

11201

Alexander E. Drilon, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 18, 2015 6:18 PM

Was further molecular profiling performed on the patient's tumor sample to identify EML4 as the upstream partner of the ALK fusion? I am guessing that this is the case. In and of itself, standard FISH cannot typically distinguish EML4-ALK from other possible rearrangement events assuming these likewise result in split or isolated signals.

While this would probably be an exceedingly rare situation (and imagining that IHC and morphology were not helpful at all), finding an upstream gene partner that has only been identified in a non-NSCLC malignancy might suggest (but not prove) that we may be dealing with a different tumor type. For example, TPM4-ALK has only been reported to my knowledge in esophageal cancer and inflammatory myofibroblastic tumor of the lung and not in NSCLC.  Similar to EML-ALK, TPM4-ALK would produce a positive FISH result, and only more comprehensive profiling would be poised to identify TPM4 as the upstream gene.

Valentina Nardi
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 19, 2015 8:39 AM

Course Faculty Response

Dr. Drilon raises a good point. ALK break apart FISH was used, which indicates an ALK rearrangement, but not which fusion partner ALK is rearranged to. That said, most EML4-ALK rearrangement being the result of an intra-chromosomal event, result in a very unique pattern of FISH break-apart signals, which can be very subtle to recognize, in that they are consistently just over two probes size apart.

To know the exact fusion partner one would have to use PCR or a dual color/dual fusion FISH assay or targeted RNA sequencing.

 

11226

Valentina Nardi
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 21, 2015 8:55 AM

Course Faculty Response

  1. As for case #1, the differential diagnosis of a CK7+ CK20- CUP includes: Lung adenocarcinoma, thyroid, salivary gland, pancreatobiliary, gastric, esophageal. The negative staining for TTF1 and the absence of a lung lesion are against a lung primary. The scattered staining for CD117 (KIT), HCG and PLAP raised the possibility of a germ cell tumor. The serum negativity for HCG and AFP and the negative subsequent immunohistochemical staining for Human placental lactogen and podoplanin were all against a diagnosis of germ cell tumor. The negative staining for WT1 and calretinin ruled out the possibility of a primary diagnosis of epithelioid mesothelioma.

 

11231

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 21, 2015 8:59 AM

Course Faculty Response

  1. Biopsy/tissue management is important in CUP and medical oncologists and pathologists should communicate regarding each patient. Small biopsy specimens are easily exhausted by multiple immunuhistochemical stains, and a small portion should be saved/reserved to perform a molecular cancer classifier assay (requires only about 500 cancer cells) in the event a single cancer type is not diagnosed by staining. Repeat biopsies should be avoided but are often necessary in CUP to make an optimal attempt at precise diagnosis. 
     
  2. The clinical setting including the site of metastasis (mediastinum), the morphology of the biopsy specimen (carcinoma), the molecular cancer classifier assay diagnosis of lung adenocarcinoma and the ALK rearrangement found by FISH testing make the diagnosis of carcinoma of the lung secure. An   anatomical   primary site (lung/bronchus or elsewhere) was not identified. Mediastinal metastasis are very common in lung cancer and, as in case #1, if the TTF-1 stain had been positive, the diagnosis of lung carcinoma would have been fairly secure  even without additional testing with the molecular cancer classifier assay. In patients with CUP who are felt likely to have cancers (lung, breast, colorectal, gastric/GE junction, melanoma) that may have highly actionable genetic alterations, it is important to obtain specific molecular testing of their cancer cells to look for known alterations such as EGFR mutations, ALK and ROS-1 rearrangements in lung cancer, Her-2-neu in breast cancer and gastric/GE junction cancers, KRAS in colorectal cancer and BRAF in melanoma.  Several other actionable genetic alterations in many solid tumors, including CUP, are likely to be relevant in the future. 

In both Cases # 1 and #2, the patients remain without any progressive carcinoma at this time. However, both will be reasonable candidates for with an ALK inhibitor therapy such as crizotinib at the time of relapse/progression.

11236

ASCO University
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 21, 2015 9:02 AM

Patient Case #2 Update

Clinical Course

Following his thoracotomy he was offered adjuvant chemotherapy for a presumed N2 metastatic adenocarcinoma of the lung, but refused any additional therapy. He has done well without clinical progression for 2.5 years. It is anticipated that should he have a recurrence, he will be a good candidate for treatment with an ALK inhibitor.

11241

Rami Manochakian, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 24, 2015 9:43 PM

Thank you for these two interesting cases. I guess I am getting to review them fairly late but I certainly enjoyed the conversation and the comments.

I do have two quick follow up management questions:

1) For case # 2, if patient agrees to receive adjuvant systemic treatment, would you consider giving Crizotinib? I know the role of adjuvant crizotinib even in typical resected lung cancer with the rearrangement is not established yet but in this particular case considering the atypical presentation and the path showing poorly differentiated carcimoma, would you consider it?  And If not using crizotinib as an adjuvant treatment, what woud be your choice of adjuvant Tx, maybe cisplatin/pemtrexed ??

2) For the same case and since we are treating it as a locally advanced stage III lung cnacer with N2 involvement, would you consider offering the patient adjuvant radiation therapy as well?

Thank you.

 

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 25, 2015 9:18 AM

Course Faculty Response

Good question. Although there are no data for ALK inhibitors in the adjuvant setting, for ALK rearranged lung cancers it appears EGFR inhibitors do not improve survival in EGFR mutated patients. The use of adjuvant crizotinib is reasonable but I would favor adjuvant chemotherapy which does improve survival in stage III adenocarcinoma. Several regimens are acceptable but I favor pemetrexed and platinum (preferably carboplatin).

11251

Valentina Nardi
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 26, 2015 10:08 AM

Course Faculty Summary

Thank you for everyone who participated in this discussion. My summary thoughts for each case are below:

  1. In a stepwise approach to a CUP case a pathologist will initially try to determine whether the cancer is a carcinoma, sarcoma, lymphoma or melanoma, both by morphological evaluation alone, for well differentiated neoplasms (case 1) or with the aid of immunohistochemical stains for poorly differentiated ones (case 2).
  2. In a carcinoma case, a wide panel of immunohistochemical stains is usually performed to narrow down the tumor subtype (adenocarcinoma, squamous cell carcinoma, neuroendocrine carcinoma, germ-cell tumor, etc.) and the primary site, which, for the most, includes at least CK7, CK20, TTF1, ER/PR (for women), PSA/PAP (for men) and other markers based on the location of the metastasis (For case 2 germ cell tumor was in the differential diagnosis of a mediastinal tumor presenting in a man).
  3. Immunohistochemical stains are not 100% specific or sensitive, as in the cases presented in which TTF1 was negative rendering the diagnosis of lung cancer less likely. A wide array of antibodies against lineage-restricted transcription factors has become available, with improved specificity for primary site determination (Conner J., Hornick Jason, Metastatic carcinoma of unknown primary: diagnostic approach using immunohistochemistry. Adv Anat Pathol 2015; 22:149-167). How many rounds of immunohistochemistry should be performed is usually decided case by case, depending on the tissue size, on the clinical presentation and upon discussion with the medical oncologists.
  4. In CUP cases in which immunohistochemical stains are inconclusive on establishing the tissue of origin, gene expression profile, as illustrated in the two cases presented, can add to the diagnostic evaluation and predict a site of origin in over 90% of cases, leading to targeted molecular genetic testing for druggable somatic alterations (in the two cases, an ALK rearrangement was detected which predicts for response to crizotinib treatment).

 

11256

F Anthony Greco, MD
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 26, 2015 10:09 AM

Course Faculty Summary

I would also like to thank everyone for participating in this discussion. Below are my thoughts based on the available information:

  1. Both the cases presented represent CUP since the anatomical primary sites were not identified.
  2. The sites of metastasis, morphology of their cancers, molecular cancer classifier assay diagnoses and ALK rearrangements make lung adenocarcinomas their very likely cancer type arising from an occult lung/bronchial primary site. 
  3. The immunohistochemical staining was consistent but not diagnostic of lung cancer or any other single cancer type. When immunohistochemical immunophenotypes are characteristic of a single cancer type a molecular cancer classifier assay is usually not needed to make a cancer type diagnosis. 
  4. The likely diagnosis of lung adenocarcinoma in both cases made by the molecular cancer classifier assay prompted testing for genetic alterations known to be present in some lung cancers and an EML4-ALK fusion was detected, further supporting the diagnoses of lung carcinoma and providing information for the possibility of precision therapy with an ALK inhibitor. 
  5. CUP patients diagnosed with specific cancer types otherwise have many of the same biologic features as their cohorts with the same easily recognized primary sites;  they respond to site specific therapy in a similar fashion and have outcomes very much like their counterparts with known advanced cancers .

11261

ASCO University
Re: Carcinoma of Unknown Primary (August 2015): Molecular Oncology Tumor Boards
Aug 26, 2015 10:11 AM

Thank you to Drs. Greco and Nardi leading the discussion of this case and also to all of those who contributed to the conversation! The forum is now closed to further comments but users have the opportunity to claim credit on ASCO University by accessing this link.

Please check back in mid-September for a new case in this series related to acute myeloid leukemia.   

 


Advertisement