The Daily Sign-Out

Proadrenomedullin in prognosis of community-acquired pneumonia

Community-acquired pneumonia is still a major healthcare problem and is under intense scrutiny as an "opportunity" to measure hospital performance (and thus determine reimbursement). With the advent of the modern antibiotic era, there has been an ongoing search for a way to assess which patients with CAP need more intensive therapy because of a greater likelihood to experience complications or death. The Pneumonia Severity Index (PSI) and CRB-65 score are established clinical scoring systems to determine mortality risk from CAP. Unfortunately, the biomarker ash heap of history is filled with biomarkers for prognosis in CAP.

I came across a fascinating article by BELLO et al. of the Hospital Universitario Miguel Servet from the May 2012 European Respiratory Journal (abstract) that adds to several previous studies on the prognostic value of proadrenomedullin (pro-ADM) in CAP.  There are now 5 studies showing that proADM improves the prognostic value of established risk scores in CAP. This follows an intriguing line of investigation moving away from the pursuit of inflammatory markers that focus on microbiological etiology toward markers that are associated with the severity of CAP and complications regardless of etiology.

 Adrenomedullin (ADM) is a member of the calcitonin peptide family. ADM is synthesized and secreted by a wide variety of cell types and synthesis is accelerated during severe infections. It exerts multiple effects including, antimicrobial activity, immune modulation by potentiating the effects of TNF-alpha, and hypotensive and vasodilatory effects. Pro-ADM is the precursor of ADM and gives rise to several biologically active products. Midregional-pro-ADM is more stable than full-length pro-ADM (and thus easier to measure) and directly reflects plasma levels of ADM.

This is a prospective cohort study of 228 immunocompetent patients with CAP. The authors measured PSI and CRB-65 scores, blood C-reactive protein, procalcitonin, and midregional (MR)-pro-ADM and performed microbiological cultures on all patients. Patients were grouped into bacterial, viral/atypical, and mixed CAP subtypes. The patients were followed at 30, 60, and 180 days and at 1 year.

RESULTS

1. Admission levels of MR-pro-ADM increased significantly with increasing severity of CAP according to both PSI and CRB-65 scores. Moreover, MR-pro-ADM was the only biomarker able to distinguish amongst all different PSI scores and was also the best biomarker for distinguishing between low-risk CAP (PSI risk classes I-III) and high-risk CAP (PSI IV-V).
2. Using an optimized cut-off of 0.646 ng/L, the sensitivity was 92%, specificity was 55%, positive predictive value was 76% and negative predictive value was 80%.
3. Multivariate analysis identified only CURB-65 and MR-proADM as predictors of complications.
4. MR-proADM was the only biomarker able to discriminate between patients who died at 30, 90 and 180 days, and 1 yr (p,0.0001).
5. No association between MR-pro-ADM and etiology was found.

There are two important conclusions from this study:

1. MR-pro-ADM is the best biomarker to predict short-term outcome in hospitalised CAP, regardless of its etiology.
2. MR-pro-ADM can also be used to monitor the long-term follow-up of CAP patients discharged from hospital. 

One caveat to this study (acknowledged by the authors) is that it is a single institution study. But this study and the four others previously reported justify serious consideration of a large multicenter prospective study to validate the prognostic value of pro-ADM in CAP.

Techniques to increase lymph node harvest from gastrointestinal cancer specimens: a systematic review and meta-analysis - Abbassi-Ghadi - Histopathology - Wiley Online Library

Good news/bad news. Another article supporting the use of fat clearing and methylene blue staining to increase lymph node yields in GI dissection BUT these techniques do not increase the identification of positive LNs or lead to upstaging.




From Evernote:




Techniques to increase lymph node harvest from gastrointestinal cancer specimens: a systematic review and meta-analysis - Abbassi-Ghadi - Histopathology - Wiley Online Library
Clipped from: http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2559.2012.04357.x/abstract;jsessionid=41CA356777E5BB95FA5D2064D29EC704.d04t02



Abstract:

Aim: This review aims to compare different histopathological techniques for lymph node harvest from ex vivo gastrointestinal cancer specimens and to examine their influence on: (i) lymph node yield; (ii) positive lymph node detection, and; (iii) cancer staging.

Method: Systematic review of the English language literature, comparing manual nodal dissection to other techniques for lymph node harvest, up to October 10th, 2011. Methodological quality of included studies was assessed.

Results: Twenty-seven studies, examining: fat clearing, methylene blue staining, fat stretching and use of a dedicated pathology assistant, were assessed. The methodological quality of the majority of included studies was poor. Meta-analysis showed that fat clearing and methylene blue staining increased mean lymph node yield, by 13 and 15 nodes respectively when compared to manual dissection. Of the 15 studies reporting positive lymph node count, two demonstrated a significant improvement for techniques other than manual dissection. Compared to manual dissection, other techniques were not shown to influence cancer staging.

Conclusion: This review has shown that fat clearing and methylene blue staining increases the mean lymph node yield from gastrointestinal cancer specimens. There is insufficient evidence to suggest that these techniques increase positive lymph node count or lead to upstaging.

© 2012 Blackwell Publishing Ltd

Implications of clonality in multifocal non small cell lung cancer

About 5% to 8% of NSCLC patients present with 2 or more anatomically separate lung tumors at diagnosis. The distinction between synchronous primary tumors and intrapulmonary metastases has important implications for staging and, thus, treatment strategies.

Arne Warth and colleagues from University Hospital Heidelberg published a single institution study in the June 1, 2012 European Respiratory Journal (abstract) analyzing the clonal relationship between multifocal NSCLC with indistinguishable morphology in a series of 78 patients.  Their analysis included LOH analysis of 14 polymorphic microsatellite markers and EGFR and KRAS mutation screening using single-strand conformation polymorphism, followed by direct sequencing of abnormally migrating bands. 

This study included 58 adenocarcinomas and 20 squamous cell carcinomas; 59 pT3, 18 pT4, and 1 pM1a. 97.4% of all specimens could be classified as clonal, likely clonal, or non-clonal; only 2 cases were non-informative.

Some results--

1. Tumor nodules of 22 patients (28%) showed divergent LOH status--indicating non-clonal origin. This is a surprising finding--remember this study only included multifocal tumors with indistinguishable morphology.
2. Activating EGFR mutations were found in both synchronous tumors in 4/58 ADC but none of the SQC.
3. KRAS mutations were found in 20 patients (17 ADC, 3 SQC) but in 6 patients the mutation was identified in only 1 of the tumors and in 1 patient there were divergent EGFR and KRAS mutations.
4. With combined allelotyping of LOH and mutational status of EGFR and KRAS, about 36% of cases (28/78 cases) showed evidence that they were not clonally related, including:
5.  14/18 ADC and 6/10 SQC located in the same lobe.
6. Kaplan-Meier curves illustrating the impact of clonality of multifocal NSCLC on overall survival did not identify a statistically significant difference but did show a tantalizing trend that patients with 2 clonally unrelated tumors have a better OS; notably, this difference did reach statistical significance by univariate analysis.

Practical "take-home" points.

There is no specific anatomical or morphological criteria that reliably indicate a clonal or nonclonal origin of multifocal NSCLC tumors. The authors reasonably recommend performing predictive molecular testing (EGFR and KRAS mutational analysis) or allelotyping separately in all cases with synchronous tumors to identify clonally related tumors. This study presents some intriguing data regarding the prognostic value of identifying nonclonalty, but the prognostic impact of clonality needs to be further analyzed in larger cohorts of NSCLC patients with multifocal synchronous tumors.

New directions for lung squamous cell carcinoma

For non-small cell lung cancer patients with localized or regional disease, treatment options vary little according to histology.  For metastatic NSCLC "targeted" therapies for patients with EGFR activating mutations or ALK translocations, pemetrexed, and the addition of bevacizumab to conventional chemotherapy have improved the treatment of metastatic NSCLC.  Unfortunately, all of these improvements apply almost exclusively to NSCLC adenocarcinoma histotype.

 For squamous cell carcinoma of the lung (SQC), the standard treatment remains cytotoxic chemotherapy, usually consisting of a platinum-containing doublet (carboplatin>US, cisplatin>Europe) with docetaxel or erlotinib used as second-line agents.

Similar to adenocarcinomas, squamous cell carcinomas are a heterogenous group of clinically, molecularly, and genetically different tumors.  Current work on SQCs is focused on identifying unique molecular patterns or driver mutations that may be therapeutic targets.  Kathryn Gold, Ignacio Wistuba and Edward Kim have published an excellent review of this recent work on SQC in Clinical Cancer Research 2012 18(11); 3002–7 that I recommend if you want a state-of-the-art update.

This review is more clinically oriented (for the pathologist) but provides connections between observed molecular abnormalities with drugs in various stages of development and current trials.  One interesting development is a phase II trial opening soon looking at the agent NCT01514864 for patients with squamous cell carcinoma with a mutation in discoidin domain receptor 2 kinase (DDR2).  I previously posted on a paper that identified DDR2 kinase mutations in lung squamous cell carcinoma about a year ago--so this is some exciting stuff! DDR2 kinase mutations occur in about 4% of SQC which doesn't seem like much until you input the denominator of the number of lung SQC diagnosed per year--this is analogous to EML-ALK translocations in lung ADC. 

Getting Ready for the ObamaCare Decision | e21 - Economic Policies for the 21st Century

Supreme Suspense: Getting Ready for the Big ObamaCare Decision | e21 - Economic Policies for the 21st Century.

This is a nicely written piece by James Capretta of the American Enterprise Institute outlining the different scenarios that might occur based on the Supreme Court decision on "ObamaCare." It is written from a conservative viewpoint but is, actually, reasonably balanced and makes some points that most everyone (right, left, and center) could agree with, at least in principle. Fundamentally, regardless of the decision, there are preexisting conditions with our health care system that must be addressed, including insurability for people with preexisting conditions, affordability of health care, providing care for the uninsured, the woeful inadequacy of the Medicaid program, the Medicare Leviathan, to name a few. 

Establishing clonality in multifocal nonsmall cell lung cancer

Up to 8% of nonsmall cell lung cancer (NSCLC) patients present with 2 or more anatomically separate lung tumors.  Distinguishing between synchronous primary tumors and intrapulmonary metastases is an important clinical question because this determination significantly changes tumor staging and subsequent therapeutic decisions.

Arne Warth et al. have published a single institution study from University Hospital Heidelberg in the current issue (1 June 2012) of European Respiratory Journal (abstract) that examines the clonal relationship in a series of patients with morphologically indistinguishable multifocal NSCLC.

The authors screened their archives for synchronous multifocal NSCLC in the same lobe (pT3), different ipsilateral lobes (pT4), and bilateral lobes (pM1a).  An important part of this study is that the authors only included those cases in which morphology between the tumors was indistinguishable using published criteria from the literature (Girard et al, Am J Surg Pathol 2009--actually a great reference).  They included 78 patients in this study.

The authors performed EGFR and KRAS mutation screening by using single-strand conformation polymorphism sensitive gel elctrophoresis and then subjected aberrently migrating bands to direct DNA Sanger sequencing.  The authors analyzed 14 polymorphic microsatellite markers for loss of heterozygosity. A clever part of the analysis of clonality in this study was the utilization of individual clonality scores for each tumor pair by multiplying the likelihood of the observed findings for each informative variable.  Thus, the clonality score defines the statistical error probability of a clonal or non-clonal relationship.

Results

1. The study included 58 adenocarcinomas (ADC) and 20 squamous cell carcinomas (SQC).  59 were pT3, 18 pT4, and 1 pM1a.  The mean follow up for all patients was 25.6 months. 97.4% of all specimens could be classified as clonal, likely clonal or nonclonal.
2. Tumors from 22 patients (28%) showed divergent LOH status indicative of non-clonal origin (14/58 ADC, 8/20 SQC).
3. Activating EGFR mutations were found in both synchronous tumors in 4/58 ADC and none of the SQC showed EGFR mutation.
4. KRAS mutations were found in 20 cases (17 ADC, 3 SQC), including in both synchronous tumors in 13 cases but in only 1 tumor in 6 cases. 1 patient had divergent EGFR and KRAS mutations in separate tumors.
5. Combined allelotyping of LOH and EGFR and KRAS mutational status showed that 36% of synchronous primary tumors were not clonally related.  Remember these were tumors that were determined to be morphologically indistinguishable.
6. Impact of clonality of multifocal NSCLC on overall survival: Kaplan-Meier curves show a trend that patients with 2 clonally unrelated tumors have a better outcome but this did not reach statistical significance by univariate analysis.

Conculsions

There is no objective, specific anatomical or morphological criteria that reliably indicate a clonal or non-clonal origin of multiple NSCLC tumors.  Therefore, the authors recommend performing predictive molecular testing (EGFR and KRAS) separately in all cases with synchronous tumor nodules or allelotyping to identifiy clonally related tumors.

This paper will influence the way I approach these situations.  First, I will specifically incorporate the Girard criteria to identify morphologically indistinguishable cases.  Second, I will send out separate tumor specimens for EGFR and KRAS mutation analysis before assigning my pathologic staging to the case.  A couple practical questions are not addressed by this study.  Given the remarkable morphological heterogeneity of lung ADC--and the high percentage of "mixed" subtype tumors, isn't it possible that morphological disparate tumors could still be clonally related?  The authors don't specifically address the issue of mixed subtype in relation to "morphologically indistinguishable" tumors--but this would be a major practical issue.  Should we simply assume that morphologically different-ish tumors are non-clonal or do the testing anyway?  A great complimentary study would test all synchronous tumors regardless of histology since this study showed that morphological criteria could not predict clonal origin even in a narrow series of synchronous tumors that looked alike in pairs.  An intriguing point of this study is the possible relationship between clonal relationship and OS.  The authors rightly conclude that this should be tested in larger cohorts of NSCLC patients to identify any statistically significance to the trend that they identified.  I am perplexed as to why clonally unrelated tumors should have better outcomes.  If any one has any thoughts, I'd love to hear them--please post a comment with your thoughts. 

Flagship Biosciences Acquires IHCtech

Flagship Biosciences released news today of their acquisition of IHCtech, a company founded by Patsy Ruegg in 2002 at the University of Colorado Health Sciences Center Pathology Department, that serves research investigators all over the U.S. with IHC services.  This is an interesting development for Flagship Biosciences and would seem to create some vertical dimension to their already formidable service line for quantitative IHC analysis.  Here's an excerpt from the press release (provided to me courtesy of Mr. Steven Potts, CEO and Chairman of Flagship):

“We have partnered with Flagship Biosciences on a number of new techniques and approaches for quantitation in immunohistochemistry,” said Patsy Ruegg, owner and founder of IHCtech. “Their approaches to whole slide analysis and commitment to quantitative pathology makes a perfect partner with IHCtech’s expertise in high quality histology and immunohistochemistry. We enthusiastically look to further innovation by evaluating all aspects of the tissue chain -- tissue procurement, fixation and processing, with the ability to measure with whole slide analysis how each of these steps contribute to variability in the overall process.”

 “While Flagship Biosciences has established strong partnerships with a number of highly respected IHC laboratories, both within the United States and internationally, this acquisition gives us the opportunity to internally evaluate tissue staining and implement new processes that better equip immunohistochemistry operations for use in quantitative pathology with whole slide imaging analysis,” said Dr. David Young, President of Flagship Biosciences. “Patsy Ruegg’s leadership in the area of high-quality research immunohistochemistry has been clear for many years. She has been a consistent expert in new IHC techniques like dual staining as well as a longtime leader in training of histology personnel.”

Usual disclaimer that I have no financial relationship etc with either company--I'm just an innocent bystander passing along information.  But I encourage you to visit Flagship's website.

 

Call It Cancer

A couple of recent editorials in The Oncologist have reacted to the recent NIH consensus panel on prostate cancer and this panel's statement that it may be "a mistake" to call low-grade, low-volume prostate cancer a cancer on the basis that it may cause unnecessary anxiety and provoke unnecessary action when the majority of men with this diagnosis are elderly and never experience morbidity or mortality with this disease.

Bruce Chabner and Matthem Smith (editorial, free registration required) opine that:

Although we agree that the connotation of lethality for some patients is unfortunate, the diagnosis of cancer depends on pathological findings (my emphasis), specifically, pathological appearance and invasive behavior. Indolent cancers occur as part of the spectrum of breast cancers, lymphoma, and other forms of malignancy, and those will continue to be called cancer.

Marshall Lichtman in another editorial draws comparison with the tortured terminology of myelodysplasic syndromes ("refractory anemia" "excess blasts") and the detrimental effects this mislabeling had on the understanding of this disease:

A profound error was made of the sort that Drs. Chabner and Smith are trying to help us avoid when the clonal cytopenias and oligoblastic myelogenous leukemia were designated myelodysplasia. By using such anachronisms (and euphemisms) as refractory anemia, and illogical terms as “excess blasts”—as if one can have an excess of malignant cells akin to excess fat cells—a neoplasia is thereby mislabeled a dysplasia, thus setting back thinking about the disorders by 50 years.

The confusion of students trying to understand why a disease with leukemic blast cells in blood and marrow is not called leukemia and of basic scientists thinking they were studying a preneoplastic condition was entirely unnecessary, if the correct language had been chosen.

This got me thinking about a similar and frequent situation I find myself with the term "carcinoid tumor"--a term that, at a minimum obscures and at worst is just downright wrong, in terms of clearly indicating the malignant potential of tumors showing typical characteristics.  So what does atypical carcinoid tumor mean?  As awkward as it is to explain metastatic carcinoid tumor in tumor board conferences, I can only imagine how it would be explaining to a patient, especially if that patient has done some research on the Internet prior to our visit.

What we call things does indeed matter and we, as pathologists, should make every effort to use terminology that reflects the behavior of what we are trying to describe.  It is some comfort that some of our oncology colleagues recognize our contribution to the diagnostic process but we must also be accountable for being complicit in using and perpetuating confusing and abstruse language.