Halling, Kevin C.

Abstract Context. —Molecular diagnostic applications that use microarrays to analyze large numbers of genes simultaneously require high-quality mRNA. As these genome-wide expression assays become more commonly used in medical practice, pathologists and oncologists will benefit from understanding the importance of obtaining high-quality RNA in order to generate reliable diagnostic and prognostic information, especially as these relate to cancer. Objective. —To review the effects that different tissue preservation techniques have on RNA quality and to provide practical advice on changes in tissue acquisition and handling that may soon be needed for certain clinical situations. Data Sources. —A review of recent literature on RNA quality, tissue fixation, cancer diagnosis, and gene expression analysis. Conclusions. —Studies have consistently shown that frozen tissue yields more intact RNA than formalin-fixed, paraffin-embedded tissue. The chemical modification, cross-linking, and fragmentation caused by formalin fixation often render RNA unsuitable for microarray analysis. Thus, when expression analysis involving hundreds or more than 1000 gene markers is contemplated, pathologists should consider freezing a specimen within half an hour (preferably within minutes) of surgical resection and storing it at −80°C or below. In coming years, pathologists will need to work closely with oncologists and other clinicians to determine when saving frozen tissue for microarray expression analysis is both practical and necessary. In select cases, the benefit of implementing a few extra tissue-handling steps may improve diagnostic and prognostic capability.

Abstract Desmoplastic infantile astrocytoma (DIA) and desmoplastic infantile ganglioglioma (DIG) are benign glioneuronal tumors that typically occur in infants, involve the superficial cerebral cortex, and have an excellent prognosis. DIA/DIG are a distinct molecular entity based on DNA methylation profiling. BRAF600 mutations are frequently reported in DIG/DIA. A recent comprehensive genetic analysis of infantile hemispheric gliomas identified 2 unique groups: group 1 harbored alterations in the receptor tyrosine kinase (RTK) genes ALK, ROS1, NTRK, and MET and group 2 harbored alterations in the RAS/MAPK pathway. We report a case of a 6.5-year-old girl who presented with seizures and right homonymous hemianopia. MRI of her brain demonstrated a large cystic/solid left hemispheric mass with remodelling of the overlying skull, consistent with a long-standing process. She underwent a gross total resection (GTR) and pathology demonstrated a DIG with a PPP1CB-ALK gene fusion (exon 5 to exon 20) identified by RNA sequencing. She remains disease free 12 months following GTR. A literature review identified 4 reported cases of pediatric brain tumors with PPP1CB-ALK gene fusions including: a 3-month-old with a hemispheric high-grade glioma which recurred 4 years later and pathology showed mature ganglioglioma, with both tumors showing the identical PPP1CB-ALK gene fusion; a 10-month-old infant with a hemispheric low-grade glioma; an infant with a “congenital” hemispheric high-grade glioma; and a child with an astrocytoma with no further clinical data. PPP1CB-ALK gene fusion appears to be a rare oncogenic driver in gliomas of infancy, including DIG.

Only 50% to 70% of patients with mesothelioma report asbestos exposure. Other exposures (eg, radiation) play a role in some cases, but some patients have no obvious cause. We describe a series of patients with long-standing indwelling intra-abdominal shunt catheters who developed malignant peritoneal mesothelioma, suggesting a novel association. We identified 7 patients who had shunts and subsequently developed mesothelioma (5 women; median age: 31 y, range: 18 to 45 y). Clinical history and pathology materials were reviewed, and RNA sequencing was performed. Clinical presentations varied; 6 patients had hydrocephalus and a ventriculoperitoneal shunt, and 1 patient had portal hypertension and a portoatrial shunt. The median duration of shunt therapy in 5 cases was 29 years (range: 12 to 35 y); the remaining 2 patients also had shunts for many years, but specific details were unavailable. Two patients had radiotherapy for malignancies in childhood. One had an alleged exposure to asbestos and 1 had prior exposure to talc. The rest had no known risk factors. Histologically, all tumors were purely epithelioid. Treatments included surgical debulking, chemotherapy, and palliative care. All 7 died of disease (median survival: 7 mo, range: 1 to 18 mo). Molecular testing showed loss of NF2 and CDKN2A/B and a BAP1 mutation in 1 case, and no genomic alterations associated with mesothelioma in 2 cases. Peritoneal mesothelioma may represent a complication of long-standing indwelling shunt catheters. The mechanism is unknown, but chronic peritoneal irritation may play a role. Albeit rare, mesothelioma should be considered in patients with a shunt who present with new ascites.

AbstractObjectivesFailure to produce sufficient quantities of functional α1-antitrypsin (AAT) can result in AAT deficiency (AATD) and significant comorbidities. Laboratory testing plays a vital role in AATD, with diagnosis requiring documentation of both a low AAT level and a mutated allele. This retrospective evaluation examines the efficacy of a liquid chromatography–tandem mass spectrometry (LC-MS/MS) (proteotyping)–based algorithm for AATD detection.MethodsA 16-month retrospective data analysis was performed on two cohorts: 5,474 samples tested with the proteotype-based algorithm and 16,147 samples directly tested by isoelectric focusing (IEF) phenotyping.ResultsLC-MS/MS reduced the rate of IEF testing by 97%. The 3% of cases reflexed to IEF resulted in 12 (0.2%) additional phenotype findings. Retrospectively applying the proteotype-based algorithm to the IEF cohort demonstrated a 99.9% sensitivity for the detection of deficiency-associated phenotypes. Most deficiency phenotypes missed by the proteotyping algorithm would come from heterozygous patients with an F, I, or P paired to an S or Z. In all of these cases, patient AAT levels were greater than 70 mg/dL, above the threshold for AAT augmentation therapy.ConclusionsThe proteotype algorithm is a sensitive and cost-effective approach for the diagnosis of clinical AAT deficiency.

AimsRecently, a novel isoform of anaplastic lymphoma kinase, with alternative transcription initiation (ALKATI), has been described in melanoma and is susceptible to targeted ALK‐inhibitor therapy. Clinical outcomes of patients with ALKATI mutated melanoma as well as correlation with immunohistochemical (IHC) methods have not yet been described.Methods and resultsClinicopathological characteristics were abstracted for 324 patients with metastatic melanoma (MM). IHC, fluorescence in‐situ hybridisation and RNA‐based digital molecular analysis assays were performed on archival tissue from 173 stage III and 192 stage IV tumours. ALKATI was identified in 12.7 and 4.8% stage III and IV tumours, respectively. Discrete presentations of the ALKATI are seen: isolated ALKATI (n = 20) and mixed ALKATI (combined ALKATI and ALKWT; n = 7). Isolated ALKWT expression (n = 4) was seen with no ALK fusions. Stage III patients showed improved survival with ALKATI expression compared to those with ALKWT or no expression [5‐year survival 80, 95% confidence interval (CI) = 57–100% versus 43%, 95% CI = 34–55%, P = 0.013]. Clinicopathological characteristics were not statistically significant. Strong diffuse cytoplasmic staining of ALK IHC (n = 12) has a sensitivity of 52.2%, specificity 100%, PPV of 100% and NPV of 92.5% of detecting isolated ALKATI.ConclusionPresence of ALKATI is a good prognostic indicator in MM. ALK IHC and digital molecular analysis can be incorporated into MM evaluation to identify patients with ALKATI for targeted therapy.

1083 Background: Pixantrone (Pix) is a novel aza-anthracenedione with structural similarities to mitoxantrone and promising activity against non-Hodgkin’s lymphoma. Due to the lack of iron binding it is theorized to exhibit less cardiotoxicity than the anthracyclines. Methods: N1031 is a phase II RCT of 2 schedules of Pix, in pts with MBC. Group A pts received 180mg/m2 IV q3 wks, and group B pts 85mg/m2 IV on days 1, 8, 15 q4 wks. Eligibility included prior exposure to anthracyclines and/or taxanes, and 1 to 3 regiments in the metastatic setting (minimum of 2 if no prior adjuvant therapy given). Due to lack of long term cardiac safety data no more than 12 cycles were allowed. Frequent cardiac imaging was performed per protocol. Primary endpoint was RR and secondary endpoints included PFS, OS, safety, and QOL. Planned sample size was 25 pts per group. Results: In total 46 pts were evaluable (23 per group), mean age 55.5 yrs (range 38-79), 37% PS 0, 52% PS 1, and 11% PS 2. 80% of pts had prior exposure to doxorubicin, 72% had prior (neo)-adjuvant therapy, 76% were ER+ and 57% received prior HT. Number of prior metastatic regiments was: 1 (28%), 2 (61%) and 3 (11%). Most common adverse events (%) of any grade were: alopecia (74), anemia (74), fatigue (85), nausea (67), ANC decrease (87), and skin disorder (41). Grade 3-4 adverse events (%) at least possibly attributed to Pix and occurring in at least 2 pts were: ANC decrease (57), fatigue (9), increased AST (4%). One pt from each group (4%) had a grade 3 decrease in EF. There were no major differences between the two groups except for more oral mucositis in group A (35% vs 4%). Median number of cycles was 3 in group A (range 1-12) and 2 in group B (range 1-8). There was only 1 confirmed tumor response per group (4%,95% CI: 0.1-22%) prompting early termination of the trial. The median PFS was 2.7 mo (95% CI: 1.8-3.8), and the median OS was 8.9 mo (95% CI: 7.5-N/A). Conclusions: Pixantrone has insufficient activity in patients with MBC exposed to prior anthracyclines and/or taxanes. Adverse events were similar to prior experience with Pix. There were no major differences between the 2 schedules of administration. There was no significant cardiac toxicity seen in this trial. Correlative studies are underway.

10019 Background: Colon cancers with microsatellite instability (MSI) display resistance to 5-fluorouracil (5-FU), and in vitro resistance can be reversed by restoring DNA mismatch repair proficiency. 5-FU inhibits the thymidylate synthase (TS) enzyme and TS may predict clinical outcome after 5-FU-based chemotherapy. To define molecular predictors of prognosis, we analyzed TS, p53, chromosome 17p allelic imbalance (AI), and patient survival stratified by MSI status. Methods: Primary colon carcinomas from patients enrolled in five 5-FU-based adjuvant therapy trials were analyzed for MSI and 17p AI using 11 microsatellite markers (MSI-H: ≥ 30% of the loci demonstrating instability). For 17p AI, markers included D17S261 and TP53 at or near the p53 locus. Expression of DNA mismatch repair (hMLH1, PharMingen; hMSH2; Oncogene), TS (TS106, Zymed), and p53 (D07, Novacastra) proteins were analyzed by immunohistochemistry. Correlations between markers and associations with overall survival (OS) were determined. Patients were censored at 5 years for DFS and at 8 years post study randomization for overall survival (OS) data. Results: Of 320 Dukes’ stage B2 and C cancers studied, 60 of 320 (19%) were MSI-H. TS expression variables (intensity, extent, weighted score) were similar in MSI-H and MSI stable/low frequency (MSS/MSI-L) cancers; similar results were found using DNA mismatch repair (dMMR) proteins. MSI-H tumors had lower stage (p= 0.0007), fewer metastatic lymph nodes (p= 0.004), and improved OS (vs. MSS/MSI-L tumors; p= 0.01). Loss of dMMR proteins was also associated with better OS (p= 0.006). None of the TS variables were prognostic for OS. Histologic grade (p= 0.0008) and nodal status (p= 0.0002) were associated with OS in contrast to 17p LOH or p53. Only MSI status or dMMR, histologic grade, and tumor stage were independent markers for OS. Conclusions: MSI-H tumors show earlier stage at presentation and better stage-adjusted survival rates. MSI status and TS expression were unrelated and TS was not prognostic, suggesting that TS levels cannot explain therapeutic resistance to 5-FU reported in MSI-H colon cancers. [Table: see text]