The Pathology Value Chain and Global Health, part 4

In this last part of our four-part series on pathology value chain, where we are using the patient’s best outcome as the maximized value, we examine two areas: Marketing/Sales and Service. The former has inherent challenges, some of which were mentioned in the last blog on outbound logistics. The latter is becoming an increasingly important component of oncology care for which many pathology labs are grasping for solutions.

In traditional business budgeting, the first step is for the marketing and sales department of a firm to provide a projection of revenue for a given period based on their knowledge of trends, markets, prior years, competition, competitive advantage, etc. These projections are then paired with costing exercises to shoot for a margin of profit. If we are going to sell $1,000,000 in widgets and it costs us $750,000 in total to make those widgets available to our customers (including costs of goods sold, administrative expense, taxes, and interest), we would have a $250,000 profit to use as retained equity or to distribute to our shareholders. When we look at pathology services for cancer, a new laboratory with no prior history may find this process extremely challenging without an enormous amount of data. An existing laboratory with many years of work may have a much easier time and, short of drastic changes in supply prices, inflation, and taxes, could likely use a simple percentage growth approach for this calculation.

But unlike widgets or iPhones or Quarter Pounders or golf clubs, no one wants to have a tissue biopsy and certainly no one wants to have suspected cancer. If we turn to epidemiological data, we can predict (and do so below) the expected number of patients in a given population to likely have cancer in the coming year (although this is clearly not the only data point we need). For a new laboratory in a place where there are no other laboratories (e.g., a small low- to middle-income country with a new Ministry of Health mandate to fight cancer), such an estimate is important for determining both if we should even have a lab (or use a regional approach) and, if we do have a lab, what our maximum volume would be assuming 100% access. The former part has been addressed previously such that there is a threshold below which is difficult to justify a lab because of the cost per sample. The latter part, however, is crucial because a “marketing campaign” (i.e., patient education and clinician education about cancer, how to diagnosis it, and how labs are part of this process) is the only way to have any volume in this laboratory.

We would except it to start slow and build but we have a finite endpoint for cancer cases in mind. But note, importantly, that the marketing campaign described has nothing to do with the pathology laboratory itself. In an existing, highly-developed market (e.g., Boston, London, Montreal, Sydney), there is a population that we can assume represents our cancer risk pool but there are also many competing laboratories (and health systems), transient use of services (e.g., Ms. Smith from Iowa decides to go to Boston for cancer care), and levels of care (i.e., low-stage cancer care in a community setting versus later-stage cancer with comorbidities in a tertiary care setting). None of these things can a given pathology laboratory control if they are in that market, but must they use all of this information to understand the projected revenue and create their budget? Or can they just assume a percentage increase? From the patient perspective, all of this is irrelevant because patients most commonly do not choose the pathology laboratory that is going to see their biopsy as it is a function of the health system to which they subscribe for their care. In that context, marketing and sales for cancer diagnostic services is largely a negotiation between laboratories and clients (e.g., clinicians, hospitals, health plans) which is often contractual. Such contracts are difficult to negotiate, take a long time, and usually last for an extended period like 1 year or longer. This very concept is contrary to the activities of the marketing and sales department which must constantly pivot, update, and change their strategy to achieve their projected revenue. It is worth noting that in many poorly developed cancer systems, patients do directly take their samples to pathology laboratories of their choice and examples of systems with kick backs to shift these samples away from government laboratories toward private practice facilities (at a much higher cost to patients) are well documented.

In the Value Chain model, service is the after-market activities of a firm to maintain their product(s) for a customer, create customer loyalty and resales, and enhance their competitive advantage through maximized firm-customer relationships. The popularity of subscription services (e.g., Amazon Prime, Netflix, Massage Envy, car leasing) stems from the increased opportunity to interact with customers continuously in low-cost ways that enhance the customer’s experience with the firm. Although a service like rending a definitive pathological diagnosis may appear to be a one-time event, recent evolution in the practice of oncology and increasing research needs have created unique servicing opportunities for pathology laboratories. The emergence of biomarkers that dictate treatment unrelated to the diagnostic process has created gaps in quality due to inefficient systems, entry cost barriers, volume challenges, and intellectual disconnect from the traditional diagnostic process. However, streamlining the biomarker process, for example, can create a competitive advantage for a laboratory and improve client loyalty and rapport.

Marketing and Sales

This activity focuses on “strategies to enhance visibility and target appropriate customers.” This activity in diagnostic anatomic pathology specifically for cancer speaks to the first part of the value chain for the patient; namely, the timely presentation of a patient to the clinical system for evaluation of cancer at the earliest possible time. As such, whether a patient presents incredibly early or very late makes no difference to the pathology laboratory because the customer choosing the pathology service is either an independent clinician or a health system. Private practice pathologists may advertise or market to community hospitals or hospital systems in hopes of capturing their volume (and revenue). Marketing for second opinion review by a pathologist can also occur and may be directly to patients. This activity is challenged from the beginning, however, due to the small market. For every 1,000,000 patients in the United States, there are about 5500 cancers per year. Assuming the accuracy of a clinical decision to obtain a biopsy is around 50% (i.e., the “malignancy rate” – when a clinician decides a biopsy is needed for suspected tumor, 50% of the time it is cancer and 50% of the time it is not), that’s 11,000 suspected cancer biopsies per million per year. Extrapolating to the US population, we get 3.6 million biopsies per year. Given that there are ~10,000 anatomic pathologists, that equates to, on average, 361 biopsies per year per pathologist (or, roughly 1 per day). Since most pathologists could easily sign out 20 cases every other day working Monday – Friday with 4 weeks of vacation annually, that’s a ratio of 1:8 (average:capacity).

The point of all of this math is that the volume of pathology work in the US that is for cancer is small relative to the total biopsies performed (or capable of being performed) by the pathology community and, thus, the market for cancer diagnostic services appears saturated. We can adjust the dial of this to take the malignancy rate to 5% (i.e., massive over biopsy setting), and find that pathology would be overwhelmed at 130% capacity just for suspected cancers; however, as we move back towards 50% malignancy rate, the average capacity is around 25% for volume. If we move on the other side of 50% towards lower biopsy rates or “improved clinical acumen,” capacity quickly drops to below 9% with a great excess of pathologists. With the promise of artificial intelligence to assist pathologists in faster sign out of higher volumes, the capacity for cancer diagnosis increases possibly 10-fold. But if you ask your average pathologist if they are busy, they report that they are. This is because the pathology laboratory, as all laboratorians are aware, processes more than just suspected cancer biopsies. Medical kidney, medical dermatology, screening colonoscopy, colposcopy, breast core needles, melanotic and non-melanotic skin lesions create a huge portion of the volume that is not part of the specific calculation above that adds many millions more samples per year to the pathology revenue stream. One framing of this case pool is that cancer biopsies, because they aren’t technically elective, are cross subsidized by providing all of the other services which are equally billable. However, this large bulk of cases are still not through direct marketing to the patient but rather to providers or health systems.

As we turn this activity towards LMICs, we instantly have a problem. There is no system in most places to support routine services for medical kidney, medical dermatology, screening colonoscopy, colposcopy, breast core needles, melanotic and non-melanotic skin lesions (especially in Black patient populations for the last). Without the cross-subsidization that these billable biopsies bring in, pathology laboratories are left with the low volumes of suspected cancer cases. As mentioned above, these laboratories are often overwhelmed to begin with so the marketing and sales activity, which would theoretically increase volume, is likely not to be a priority. In these settings, however, what will increase volume and improve the quality of care for patients is large pre-analytical efforts by governments and other entities to educate the public and the general practitioner about cancer screening and diagnosis, community awareness about cancer care systems, specimen transport networks from the most rural directly to pathology laboratories, and government spending on prevention of cancer.

Service

This last set of activities are to “maintain products and enhance consumer experience.” For a diagnosis of cancer, once rendered, there are many potential touch points with both the patient and the treating clinician that can enhance the outcomes for the patient. These include maintenance of tissue in repositories for future studies, performance of future studies related to newly available treatments, access to clinical trials, and, as mentioned in the outgoing logistics, increased, and enhanced communications around the diagnosis and subsequent information. In LMICs, there is a great desire to provide such enhancements especially in settings where these activities can facilitate local research and generate much-needed local clinical trials with pharmaceutical and other industry partners. As the other steps of the value chain are improved, the continue service will come into focus and can include such activities as external quality assurance, laboratory accreditation, personnel certification, documented compliance with standards, awards, and other accolades.

To conclude, from the patient framework, the maximum value for a patient with cancer involves the earliest possible detection of the tumor and a rapid, accurate diagnostic report matched to treatment options that lead to survivorship. For a pathology laboratory, the best outcomes for patients and the best revenue model for the laboratory results from a high-volume of small samples (i.e., biopsies) reported with complete clarity. Cross subsidization of cancer diagnostic services (especially those for later staged, complex cancer patients) with other non-cancer, pathology-based reporting is crucial to create a sustainable revenue stream and ensure highest quality outcomes. Competitive advantage in pathology services specific to cancer are currently and will continued to be largely tied to the after diagnostic service and support to keep the patient on the most beneficial cancer journey.

References

  1. Porter, M. (1985). The value chain and competitive advantage, Chapter 2 in Competitive Advantage: Creating and Sustaining Superior Performance. Free Press, New York, 33-61.
  2. Histology. Wikipedia. https://en.wikipedia.org/wiki/Histology#:~:text=In%20the%2019th%20century%20histology,by%20Karl%20Meyer%20in%201819.
  3. Thorpe A et al. The healthcare diagnostics value game. KPMG International. Global Strategy Group. https://assets.kpmg/content/dam/kpmg/xx/pdf/2018/07/the-healthcare-diagnostics-value-game.pdf
  4. Digital Pathology Market CAGR, Value Chain Study, PESTEL Analysis and SWOT Study|Omnyx LLC, 3DHISTECH Ltd, Definiens AG. https://www.pharmiweb.com/press-release/2020-06-30/digital-pathology-market-cagr-value-chain-study-pestel-analysis-and-swot-study-omnyx-llc-3dhistec
  5. Friedman B. The Three Key Components of the Diagnostic Value Chain. Lab Soft News. January 2007. https://labsoftnews.typepad.com/lab_soft_news/2007/01/the_three_eleme.html
  6. XIFIN. The Evolution of Diagnostics: Climbing the Value Chain. January 2020. https://www.xifin.com/resources/blog/202001/evolution-diagnostics-climbing-value-chain
  7. Sommer R. Profiting from Diagnostic Laboratories. November 2011. Seeking alpha. https://seekingalpha.com/article/305931-profiting-from-diagnostic-laboratories#:~:text=The%20three%20year%20average%20operating,current%20operating%20margin%20of%2012.9%25.
milner-small


-Dan Milner, MD, MSc, spent 10 years at Harvard where he taught pathology, microbiology, and infectious disease. He began working in Africa in 1997 as a medical student and has built an international reputation as an expert in cerebral malaria. In his current role as Chief Medical officer of ASCP, he leads all PEPFAR activities as well as the Partners for Cancer Diagnosis and Treatment in Africa Initiative.

Microbiology Case Study: Blood stream infection in a 77 year old patient – is it Really from Mosquitoes?

A 77 year old male with a past medical history of end stage renal disease (ESRD) on hemodialysis, type 2 diabetes, coronary and peripheral artery disease, and squamous cell carcinoma of the lung on current chemotherapy/radiation was admitted to our hospital from his outpatient hematology oncology clinic for acute hypoxia. Due to an episode of decreased responsiveness and a potential stroke, Head computed tomography (CT) and computed tomography angiography (CTA) were performed. Electroencephalography showed diffuse slowing, suggestive of encephalopathy. Three days after admission, he became hypotensive and febrile. Pulmonology/critical care was consulted; blood and respiratory samples collected for cultures. The blood culture grew gram negative rods in the aerobic bottles (Images 1-3) after overnight incubation. The patient was initially on cefepime and switched to meropenem 500 mg IV daily.

The following day, the blood culture isolate was identified as Elizabethkingia anophelis. The isolate was resistant to both of the patient’s prior inpatient antibiotics, cefepime and meropenem. Additionally, the isolate was resistant to first, second, and third generations of cephalosporins, aztreonam, tetracyclines and tobramycin. However, it was susceptible to amikacin, ciprofloxacin, gentamicin, and trimethoprim/sulfamethoxazole. Meropenem was discontinued and replaced with ciprofloxacin 400 mg IV daily. Infectious disease was consulted; at this time the patient was displaying nuchal rigidity and extreme encephalopathy. Increased dosing of Ciprofloxacin for better central nervous system penetration, in combination with trimethoprim/sulfamethoxazole 2.5mg/kg IV q8h, rifampin 600 mg IV daily was recommended and a lumbar puncture to be performed once the patient was stable. Sadly, due to underlying severe comorbidities, along with worsening CNS responses, the patient expired on day 9.

Image 1. Small gram negative rods of E. anophelis from positive aerobic blood cultures.
Image 2. Blood Agar Plate growing E. anophelis after overnight incubation at 35 degrees C
Image 3: Chocolate plate growing smooth creamy gram negative E. anophelis after overnight incubation at 35 degrees C

Discussion

The genus Elizabethkingia was named after Elizabeth O. King, a microbiologist at the Center for Disease Control (CDC) and Prevention, who discovered many medically important bacteria in the late 1940s to early 1960s. This included describing Elizabethkingia meningoseptica (formerly Chryseobacterium meningosepticum) in 1959. Elizabethkingia and Kingella genera, and the species Kingella kingae are also named in her honor.1 Elizabethkingia is a gram negative, obligate aerobic bacillus. It was classified under the families Flavobacteriaceae and Chryseobacterium, but was reclassified as Elizabethkingia in 2005.2 E. meningoseptica, the most frequently isolated Elizabethkingia species, has been implicated in cases of neonatal sepsis, meningitis, and nosocomial pneumonia.3

On the other hand, E. anophelis was recently characterized in 2011 and was initially thought to be underrepresented likely due to the genotypic and phenotypic similarity to E. meningoseptica.4,5,6 The species name Elizabethkingia anophelis originated from the anopheles mosquito as it has been isolated from the midgut of Anopheles gambiae mosquitoes.4 The role of mosquitos in maintenance and transmission of Elizabethkingia anophelis is unclear.4,6 Oxidase and catalase positive E. anophelis, (Fig 1) grows well on blood and chocolate agar plates (Fig 2 and 3) as smooth and slight-yellowish colonies although it does not grow on MacConkey agar.7

Beginning late 2015, an increased number of Elizabethkingia infections were identified in Southeastern Wisconsin. Between November 2015 to May 2016, 63 cases of E. anophelis were reported to the Wisconsin Division of Public Health. Cases spread across Illinois and Michigan as well, making it the largest E. anophelis outbreak described to date. A case series published from Froedtert Health System hospitals described their experience with E. anopheles.8 This was a retrospective case series of all consecutive patients admitted to Froedtert Health System hospitals with positive cultures of any site for Elizabethkingia, Flavobacterium, and Chryseobacterium from November 2015 to June 2016. In this time period, 11 patients were identified with cultures positive for E. anophelis. All patients had positive blood cultures at the time of hospital admission. E. anophelis was identified in both sterile and nonsterile body fluids. All 11 patients had at least one major comorbidity, including cancer, COPD, diabetes, ESRD requiring hemodialysis, and alcohol abuse. Two patients died within 30 days of a positive E. anopheles culture (attributable mortality rate, 2/11 = 18.2%).5, 8, 9

Interestingly, vertical transmission of E. anopheles causing neonatal meningitis has been reported.6 Molecular evidence suggested vertical transmission from a mother with chorioamnionitis, but a mechanism of colonization for the mother could not be found and environmental contamination was not also found.6,7 On the other hand, taps and aerators contaminated with E. anophelis in an intensive care unit has been reported.10 E. anophelis should be treated as a true pathogen, particularly in patients with multiple comorbidities.8 Isolation in sterile fluid should never be considered a contaminant.

Since Elizabethkingia is a non-glucose, non-lactose-fermenter, the MIC breakpoint of E. anophelis is reported based on those of non-Enterobacterales Table 2B-5 of CLSI (Clinical Laboratory Standard Institute) M100 guidelines. Elizabethkingia species, including E. anophelis, are intrinsically resistant to several antibiotics and produce elevated MIC on in vitro susceptibility tests. A number of species also harbor beta-lactamase/metallo beta-lactamase (MBL) genes. Empirical treatment should include piperacillin/tazobactam plus quinolone, rifampin, or minocycline. Vancomycin has been used in severe infections, especially meningitis. The best duration of therapy has not been evaluated by clinical trials8.

In summary, our patient acquired this infection in the setting of multiple chronic comorbidities and was immunocompromised due to active malignancy and recent chemotherapy. He has a similar clinical profile to the other patients in the above-mentioned study. One notable difference is that our patient’s isolate was resistant to cefepime, where the isolates from this outbreak were susceptible. After discussion with our infectious disease colleagues regarding this case, we agreed his cause of death was likely multifactorial, though this infection may have been a significant contributing factor.

References

  1. KING EO. Studies on a group of previously unclassified bacteria associated with meningitis in infants. Am J Clin Pathol. 1959 Mar;31(3):241-7. doi: 10.1093/ajcp/31.3.241. PMID: 13637033.
  2. Kim KK, Kim MK, Lim JH, Park HY, Lee ST. Transfer of Chryseobacterium meningosepticum and Chryseobacterium miricola to Elizabethkingia gen. nov. as Elizabethkingia meningoseptica comb. nov. and Elizabethkingia miricola comb. nov. Int J Syst Evol Microbiol. 2005 May;55(Pt 3):1287-1293. doi: 10.1099/ijs.0.63541-0. PMID: 15879269.
  3. Jean SS, Lee WS, Chen FL, et al. Elizabethkingia meningoseptica: an important emerging pathogen causing healthcare-associated infections. J Hosp Infect 2014; 86:244–9.
  4. Kämpfer P, Matthews H, Glaeser SP et al. . Elizabethkingia anophelis sp. nov., isolated from the midgut of the mosquito Anopheles gambiae. Int J Syst Evol Microbiol 2011; 61(Pt 11):2670–5. [PubMed] [Google Scholar]
  5. Perrin A, Larsonneur E, Nicholson AC, et al. Evolutionary dynamics and genomic features of the Elizabethkingia anophelis 2015 to 2016 Wisconsin outbreak strain. Nat Commun 2017; 8:15483.
  6. Lau, Susanna K.P.; Wu, Alan K.L.; Teng, Jade L.L.; Tse, Herman; Curreem, Shirly O.T.; Tsui, Stephen K.W.; et al. (February 2015). “Evidence for Elizabethkingia anophelis Transmission from Mother to Infant, Hong Kong”. Emerging Infectious Diseases. 21 (2): 232–241. doi:10.3201/eid2102.140623. PMC4313635. PMID25625669
  7. Koneman’s Color Atlas and Textbook of Diagnostic Microbiology. 7th Edition. 2016.
  8. Castro, C. E., Johnson, C., Williams, M., Vanderslik, A., Graham, M. B., Letzer, D., . . . Munoz-Price, L. S. (2017). Elizabethkingia anophelis: Clinical Experience of an Academic Health System in Southeastern Wisconsin. Open Forum Infectious Diseases, 4(4). doi:10.1093/ofid/ofx251
  9. Wisconsin Department of Health Services; Elizabethkingia 2017. Available at: https://www.dhs.wisconsin.gov/disease/elizabethkingia.htm. Accessed 9 January 2017. [Google Scholar]
  10. Balm MN, Salmon S, Jureen R, Teo C, Mahdi R, Seetoh T, Teo JT, Lin RT, Fisher DA. Bad design, bad practices, bad bugs: frustrations in controlling an outbreak of Elizabethkingia meningoseptica in intensive care units. J Hosp Infect. 2013 Oct;85(2):134-40. doi: 10.1016/j.jhin.2013.05.012. Epub 2013 Aug 17. PMID: 23958153.

-J. Stephen Stalls, MD is a PGY-II pathology resident at the East Carolina University Department of Pathology and Laboratory Medicine. He plans to pursue hematopathology and molecular pathology fellowships, but also greatly enjoys his time in the microbiology lab. Outside of work, he enjoys playing the drums and going to concerts.

-Phyu Thwe, Ph.D., D(ABMM), MLS (ASCP)CM is a Technical Director at Vidant Medical Center Clinical Microbiology Laboratory. She completed a Clinical and Public Health Microbiology Fellowship through a CPEP-accredited program at the University of Texas Medical Branch (UTMB) in Galveston, Texas. She is interested in extrapulmonary tuberculosis and developing diagnostic algorithms.

The Role of Pathology and Laboratory Services in Global Colorectal Cancer Prevention

Globally, colorectal cancer (CRC) is the third most common cancer in men and the second in women.1 It is the fourth main cause of cancer death in the world, with nearly 1.8 million new cases and 881,000 deaths in 2018.2 As alarming as these numbers, some progress has been made in terms of disease occurrence and outcome in many developed countries through the design and implementation of effective screening programs. With better access to healthcare services and overall improvements in treatment of CRC, patients in developed countries can have their pre-malignant, in-situ and minimally invasive polyps detected and removed in time through effective colonoscopy screens and disease interpretation by pathologists. Unfortunately, this progress is not uniform across the globe. Many developing countries across Latin America, Africa and Asia are experiencing increases in their CRC cases.3-5 A number of factors are responsible for this disparate reality.

With limited healthcare resources, many developing countries still struggle with efficient and effective health services. Several studies have shown the significant role of effective screening programs in detecting early colorectal adenomas. However, channeling scarce resources to support preventative health services is still a luxury many of these countries cannot afford. In addition, making sure these services actually work, would require effective laboratory services, laboratory professionals and pathologists. Unfortunately, due to limited resources and ambiguous priorities, laboratory services in some areas are not equipped to prioritize preventive health services, with direct impacts on CRC incidence and survival.

Image 1. Hematoxylin and Eosin-stained composite image of Medullary Colon Cancer. Left side (4x magnification) shows colonic mucosa with a well-demarcated solid nest of tumor cells with conspicuous lymphoplasmacytic infiltrates. Right side (20x magnification) shows a higher magnification of the pleomorphic tumor cells with irregular nuclear membranes, vesicular chromatin, prominent nucleoli and multiple mitotic figures. Medullary colon cancers are usually right-sided and have a better prognosis compared with poorly-differentiated or undifferentiated adenocarcinoma of the colon.6 

Even though the majority of CRC occur through somatic events, some however, do progress through well-defined germline mutations including inherited cancer syndromes including Lynch syndrome (Hereditary Non-Polyposis colon cancer/HNPCC), Peutz-Jeghers syndrome and the Familial Adenomatous Polyposis (APC mutations) pathway. Unfortunately, cancer genetics and molecular diagnostics is still not mainstream in healthcare institutions in many developing countries. Therefore, patients and their families with affected mutations may find it extremely difficult getting access to the care they need in terms of diagnosis and treatment.

The rising incidence of CRC in developing countries may also be explained by the rising trends in Westernized practices which leads to several modifiable risk factors including the consumption of diets rich in saturated fats, lack of physical activity, diabetes, obesity, alcohol consumption and smoking. Preventive health services through effective public health education on the dangers and risks of these environmental practices may play a role in disease prevention and outcomes.

At the crux of CRC prevention and early detection is effective screening programs. As March marks colorectal cancer awareness month, it is imperative to emphasize that any sustainable health policy program must consider the unique role that effective pathology and laboratory services has to play. We must be invited to stakeholder discussions as the value we bring to such discussions cannot be overstated. A failure to recognize our position as central to improving patient outcomes has made many healthcare systems less effective in addressing public health challenges.

References

  1. GLOBOCAN. Estimated cancer incidence, mortality and prevalence worldwide in 2012. 2012. http://globocan.iarc.fr/Default.aspx
  2. Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68:394–424.
  3. Bosetti C, Malvezzi M, Chatenoud L, et al. Trends in colorectal cancer mortality in Japan, 1970‐2000. Int J Cancer 2005;113:339–41.
  4. Center MM, Jemal A, Ward E. International trends in colorectal cancer incidence rates. Cancer Epidemiol Biomarkers Prev 2009;18:1688–94.
  5. Souza DL, Jerez‐Roig J, Cabral FJ, et al. Colorectal cancer mortality in Brazil: predictions until the year 2025 and cancer control implications. Dis Colon Rectum 2014;57:1082–9.
  6. Cunningham J, Kantekure K, Saif MW. Medullary carcinoma of the colon: a case series and review of the literature. In Vivo. 2014;28(3):311-314.

-Evi Abada, MD, MS is a Resident Physician in anatomic and clinical pathology at the Wayne State University School of Medicine/Detroit Medical Center in Michigan. She earned her Masters of Science in International Health Policy and Management from Brandeis University in Massachusetts, and is a global health advocate. Dr. Abada has been appointed to serve on the ASCP’s Resident’s Council and was named one of ASCP’S 40 under Forty honorees for the year 2020. You can follow her on twitter @EviAbadaMD.

Hematology Case Study: CBC with >80% Blasts

The patient is a 67 year old male who first visited his dentist at the end of December complaining of pain in the jaw that he had been experiencing since early Dec. He had put off making an appointment because he didn’t want to have to go to the doctor with COVID precautions, but the pain was now radiating to his teeth, so he made a dentist appointment. The dentist found no evidence of abscess or other infection but ‘adjusted his bite’. The patient was advised to take over the counter NSAIDs as needed or pain but no prescriptions was needed. Three weeks later the patient visited an urgent care because he had no improvement of the jaw pain. At this time he relayed symptoms of cough, fever, chills, night sweats and chronic fatigue. Patient history included an active lifestyle with vigorous aerobic exercise several times a week, but the he stated that he had been feeling too fatigued to exercise for over a month. On exam the patient was found to be tachycardic with bilateral tonsillar lymphadenopathy and oropharyngeal exudate. The patient was tested for COVID, influenza and Group A Strep. The COVID-19 was negative, as was the influenza A and B, but the Group A Strep was positive. The patient was sent home with a prescription for antibiotics.

One week later, the patient called his PCP because he still had cough, fever and chills and now was experiencing shortness of breath. The office directed the patient to go to the ER but the patient was reluctant to go to the hospital and stated he would rather be seen at the office. On review of the patients chart, the PCP agreed to see him in the office because he had had a negative COVID test in the past week. Two days later the doctor examined the patient in his office and still suspected COVID-19. He ordered a PCR COVID-19 test along with CBC/differential and erythrocyte sedimentation rate (ESR). We received a routine CBC on the patient. Results are shown below.

The patient had no previous hematology or oncology history and no previous CBC received at our lab. The critical WBC was called to the physician. Based on the WBC and flags on the auto differential, a slide was made and sent to our CellaVision (CV). On opening the slide in CV, we immediately called our pathologist for a pathology review. A rare neutrophil was seen on the peripheral smear, with immature appearing monocytes, few lymphocytes and many blasts.

Image 1. Images from CellaVision.

The pathologist reviewed the slide and the sample was sent for flow cytology studies and FISH. The pathologist’s comment ”Numerous blasts (>60%) consistent with Acute Myeloid Leukemia(AML). Specimen to be submitted for flow cytometry. Hematology consult recommended” was added to the report.

Image 2. Image from CellaVision. Predominately blasts with one neutrophil seen in field of unremarkable RBCs.
Image 3. Image from CellaVision.

The myeloid/lymphoid disorders and acute leukemia analysis by flow cytometry reported myeloblasts positive for CD117,CD33, and CD13. Final interpretation was Acute Myeloid leukemia (non-M3 type).

AML is the most common form of leukemia found in adults. AML was traditionally classified into subtypes M0 through M7, based on the cell line and maturity of the cells. This was determined by how the cells looked under the microscope after a series of special staining techniques, but did not take into account prognosis. It is now known that the subtype of AML is important in helping to determine a patient’s prognosis. In 2016 World Health Organization (WHO) updated the classification system to better address prognostic factors. They divided AML into several broad groups, including AML with certain chromosomal translocations, AML related to previous cancer or cancer therapy, AML with involvement of more than one cell type, and other AML that don’t fall into the first three groups.2 Once a case has been placed in one of these broad groups, the AML can be further classified as poor risk, intermediate risk and better risk based on other test results. Better risk is associated with better response to treatments and longer survival.3 The European LeukemiaNET (ELN) first recommended integrating molecular and cytogenic data into classification to create such a risk classification system for AML in 2010 (ELN-2010). In 2017, this was again revised (ELN-2017) to further improve risk stratification. The ELN-2017 can be used to more accurately predict prognosis in newly diagnosed AML.1

What this means is that AML is now classified by abnormal cell type as well as by the cytogenetic, or chromosome, changes found in the leukemia cells. Certain chromosomal changes can be matched with the morphology of the abnormal cells. These chromosomal changes can help doctors determine the best treatment options for patients because these changes can predict how well treatment will work.

Examples of risk classification include the knowledge that some chromosome rearrangements actually offer a better prognosis. For example, a translocation between chromosomes 15 and 17 [t(15;17)] is associated with acute promyelocytic leukemia (APL or M3). APL is treated differently than other subtypes and has the best prognosis of all the AML subtypes. Other favorable chromosomal changes include [t(8;21)] and [inversion (16) or translocation t(16;16)]. Examples of intermediate risk prognosis are ones associated with normal chromosomes and [t(9;11)]. Poor prognosis is associated with findings such as deletions or extra copies of certain chromosomes or complex changes in many chromosomes.3

The patient was diagnosed with AML, non M3 type. AML prognosis is based on CBC results, markers on the leukemia cells (flow cytometry), chromosome (cytogenic) abnormalities found and gene mutations (molecular abnormalities). In this patient the FISH studies did not demonstrate any chromosome rearrangements, which alone would place him in an intermediate risk group. In addition, our patient was over age 60 and had a WBC over 100,000/mm3 which have both been linked to worse outcomes.

Here’s one more photo for your enjoyment! It’s not often that we see so many blasts in a patient with no previous history. As a side note, I was contemplating titling this blog “Fatigue and Shortness of Breath in the Time of COVID.” I can’t help but wonder if this patient would have been diagnosed 6-8 weeks earlier if this was another year and he had been seen when he first experienced symptoms. This year, emergency rooms and physicians have reported a decrease in numbers of patients being seen for chest pain, ketoacidosis, shortness of breath, strokes and other serious conditions. Many patients are reluctant or afraid of sitting in crowded waiting rooms, fearful they will catch COVID. And many doctors are only offering virtual visits or have reduced the number of patients being seen so it is harder to get appointments. This patient expressed his reluctance to seek medical help because of fears of COVID. He did not want to go out in public and waited almost a month for symptoms to go away on their own before first being seen. After going to the walk in center, he called his PCP a week later and was still averse to going to the ER as suggested by the doctor. Then he waited another 2 days for an office appointment. The doctor still suspected COVID, but fortunately for the patient, ordered a CBC. The flow cytometry and FISH studies were available the following day. The patient was referred for hematology consult but has not been seen again at our hospital.

Image 4. More images from CellaVision.

References

  1. Boddu, P.C., Kadia, T.M., Garcia‐Manero, G., Cortes, J., Alfayez, M., Borthakur, G., Konopleva, M., Jabbour, E.J., Daver, N.G., DiNardo, C.D., Naqvi, K., Yilmaz, M., Short, N.J., Pierce, S., Kantarjian, H.M. and Ravandi, F. (2019), Validation of the 2017 European LeukemiaNet classification for acute myeloid leukemia with NPM1 and FLT3‐internal tandem duplication genotypes. Cancer, 125: 1091-1100. https://doi.org/10.1002/cncr.31885
  2. Mandel, Ananya. Acute Myeloid Leukemia Classification. Medical Life Sciences. https://www.news-medical.net/health/Acute-Myeloid-Leukemia-Classification.aspx
  3. Ari VanderWalde, MD, MPH, MA, FACP; Chief Editor: Karl S Roth, MD. Genetics of Acute Myeloid Leukemia. Medscape. Updated: Dec 17, 2018 
Socha-small

-Becky Socha, MS, MLS(ASCP)CMBBCM graduated from Merrimack College in N. Andover, Massachusetts with a BS in Medical Technology and completed her MS in Clinical Laboratory Sciences at the University of Massachusetts, Lowell. She has worked as a Medical Technologist for over 40 years and has taught as an adjunct faculty member at Merrimack College, UMass Lowell and Stevenson University for over 20 years.  She has worked in all areas of the clinical laboratory, but has a special interest in Hematology and Blood Banking. She currently works at Mercy Medical Center in Baltimore, Md. When she’s not busy being a mad scientist, she can be found outside riding her bicycle.

Eye Spy 2.0

Now that you’ve seen some malignant ocular entities (both primary and metastatic), I’d like to share some (in my eyes) curveballs. Working in a cancer center, I see more cancer than I do benign processes. You can say I’m fine tuned to identifying malignant cells, and I find it difficult to rest assured that what I’m diagnosing truly is benign. When you find one malignant cell in a pleural fluid, changing the patient’s diagnosis from Stage II to Stage IV, that primed search for atypia is always on. With that said, not every specimen screened at a cancer center is malignant. We rule out malignancy and confirm benign processes as well. As for benign ocular FNAs, well… you have to be as certain calling benign as you do malignant because the process of acquiring an additional sample to confirm your diagnosis is just as involved and unpleasant as the first biopsy.

Here, I present six non-malignant eye FNA cytopreparations that had me searching for atypia longer than I expected.


Case 1. A 72 year old female with a history of breast carcinoma; She presented with a choroidal nevus, OD in 2014. The oncologist noted a slight increase in base and thickness six years after the initial evaluation. The clinical diagnosis is a choroidal nevus, rule out low-grade melanoma.

Images 1-2: Eye, Right, Choroid, Fine Needle Aspiration. Pap-stain.

Final Diagnosis: No malignant cells identified. A few benign-appearing melanocytes, consistent with nevus.

Note: Material is scant cellular and consists mainly of scattered retinal pigment epithelial cells and sensory cells. There are a few benign-appearing spindle-shaped cells with bland nuclei, consistent with nevus. However, the paucity of the sample precludes a definitive diagnosis. Recommend clinical correlation to exclude sampling error.


Case 2. A 21 month old male with a history of an iris stromal cyst, OS. Lesion is status post-FNA (twice), had regrowth, and aspirated for the third time. Previous FNAs showed no malignant cells.

Images 3-4: Eye, Left, Choroid, Fine Needle Aspiration. Pap-stain.

Final Diagnosis: No malignant cells identified. Rare pigment-laden macrophages, a few lymphocytes, and benign surface epithelial cells present. (Consistent with stromal cyst).


Case 3. A 65 year old female with a history of nevus, OD. Ocular oncology followed up and noted an increase in thickness from 2.76 to 3.00 mm. In light of family history of cutaneous melanoma, treatment is suggested. Clinical diagnosis: ciliary body melanoma.

Images 5-6: Eye, Right, Choroid, Fine Needle Aspiration. Pap-stain.

Final Diagnosis: Scant mildly atypical amelanotic spindle cells, consistent with nevus.

Note: Diagnostic material is very scant. There are a few benign-appearing single spindle cells and one cluster of cohesive spindle cells with no pigment. Prominent nucleoli are not seen. While the cytologic features in the context of clinical presentation is consistent with nevus, the paucity of the sample precludes a definitive diagnosis. There is no diagnostic evidence of malignant melanoma in this sample. Recommend clinical correlation to exclude sampling error.


Case 4. A 9 year old female with a history of retinoblastoma, OD; status post IAC (intra-arterial chemotherapy) x5, IVit melphalan x 4, and plaque radiotherapy. She recently developed vitreous hemorrhage (OD). Since the hemorrhage is obscuring the retinal view, vitrectomy is planned.

Image 7: Eye, Right, Vitreous, Fine Needle Aspiration. Pap-stain.

Final Diagnosis: No malignant cells identified. Lymphocytes and histiocytes present.


Case 5. An 11 year old female with vitreal retinoblastomas, OU; status post CRD therapy x6; plaque I-125, EBRT, proton beam; IVit Metphalan x2, and PPV (pars plana vitrectomy), OS. She now presents with dense vitreous hemorrhage blocking view since 2019.

Images 8-9: Eye, Left, Vitreous, Fine Needle Aspiration. Pap-stain.

Final Diagnosis: No malignant cells identified. Proteinaceous material and scattered small lymphocytes present.


Case 6. A 52 year old female with no cancer history. In 2019, she presented with pain, floaters, decreased visual acuity, and photophobia, OS. She was treated with antibiotic drops and steroids. In 2020, she was noted to have panuveitis with yellow iris nodules. FNA showed necrotic cells inadequate for diagnosis, as well as a negative culture and PCR for HSV, toxoplasmosis, CMV, and VZV. Clinical diagnosis: suspicious for iris lymphoma, OS.

Images 10-11: Eye, Left, Anterior Chamber, Fine Needle Aspiration. Pap-stain.

Final Diagnosis: Mixed inflammatory cells, favor an inflammatory process.

Note: There are neutrophils, lymphocytes, and scattered histiocytes in a background of numerous fragments of an amorphous substance. The cellularity is inadequate for flow cytometric analysis. We performed immunocytochemical stains on cytospin preparations. The lymphocytes are predominantly T-cells showing positive staining for CD45 and CD3 and negative for CD20. There is no diagnostic evidence of malignant lymphoma in this sample. The amorphous substance may represent lens fragments, in which case the possibility of lens-induced uveitis should be considered in the differential diagnosis. Recommend clinical correlation.


As you can tell, benign processes are complicated and require thorough explanations in our pathologic diagnoses, especially when they differ from the clinical impression. The majority of benign ocular FNAs are paucicellular, and we make the most of what we have through optimal preservation and preparation of cellular material. And don’t forget-not everything is cancer. Cyst contents, proteinaceous debris, and inflammatory cells make up a share of cases we see, and it’s okay to diagnosis them as benign. Just keep an eye out.

-Taryn Waraksa, MS, SCT(ASCP)CM, CT(IAC), has worked as a cytotechnologist at Fox Chase Cancer Center, in Philadelphia, Pennsylvania, since earning her master’s degree from Thomas Jefferson University in 2014. She is an ASCP board-certified Specialist in Cytotechnology with an additional certification by the International Academy of Cytology (IAC). She is also a 2020 ASCP 40 Under Forty Honoree.

Breakpoint Breakdown II: Breakpoints at the Bedside

A critical task for clinical microbiologists is interpreting antimicrobial susceptibility testing (AST) results for microorganisms recovered from patient specimens. But what happens to that information after it is passed along to the clinical team, and how does it influence patient care at the bedside? In our previous blog, we discussed details concerning how breakpoints are established and identified them as an indispensable component of appropriate and effective antimicrobial prescribing. Here we will discuss the more practical application of breakpoints to guide patient care. To aid in the discussion, I’ve once again recruited two infectious diseases (ID) pharmacists from the UT Southwestern Medical Center to provide their valuable prospective on the use of AST results both at the bedside and to optimize antimicrobial stewardship initiatives.

Applying Breakpoints in Clinical Care

In the most simplistic terms, the reporting of isolate’s categorical susceptibility (susceptible, intermediate, and resistant) provides guidance concerning which antimicrobials will likely be effective to treat an infection.1 Many microbiology laboratories, however, report additional AST information including minimal inhibitory concentrations (MICs) and phenotypic information concerning resistance mechanisms (e.g. production of Extended Spectrum Beta-Lactamases (ESBLs) or inducible-clindamycin resistance)2. This more detailed information is often utilized by ID specialists to help guide and optimize beside management.

Unfortunately, several misapplications of AST data can be encountered in clinical care (Image 1). One common error is simply selecting the antimicrobial with the lowest MIC for treatment. This runs contrary to the idea that each antimicrobial has an individualized breakpoint for a given pathogen or pathogen group. Therefore, a low MIC for antimicrobial “A” may be at or near the breakpoint, but higher MIC for antimicrobial “B” may be a dilution or more lower than the breakpoint. Another common misconception is that if the isolate is reported susceptible to a given antimicrobial, that antimicrobial is the optimal choice for the patient, and will work 100% of the time! Very little in medicine is 100%, and this holds true for antimicrobials. Even if the optimal agent is selected based on AST data, there is no guarantee it will be effective in the patient. Many factors influence the effectiveness of antimicrobials including patient characteristics (e.g., immune system, kidney function), drug characteristics (i.e., pharmacokinetics [PK]/pharmacodynamics[PD]), and bug characteristics (e.g., underlying resistance present despite reporting as susceptible or potential for inducible resistance).2 This illustrates why clinicians may prefer to have all available information to further aid in optimization of antimicrobial therapy beyond the information provided by categorical reporting.

But why is an MIC needed – isn’t just knowing something is susceptible or resistant enough? In particular, more granular information such as the actual MIC for a given bug-drug combination may aid in more intricate decisions surrounding the choice of agent and dose. One argument for routine reporting of MIC values is the ability to better discern how near the breakpoint the particular MIC is for a given drug. The standard MIC reporting error is plus or minus one doubling dilution. When this is factored in, it may provide additional context as to how near or over the isolates phenotype is to assigned breakpoint.2 Clinicians may further couple this with their understanding of PK/PD to determine the optimal agent based on antimicrobial properties and microbial characteristics.

A counter argument is that the MIC values are prone to variation depending on testing modality and may not represent truly what occurs in actual humans (see Mouton JW et al. for more detailed discussion).3 As mentioned above, some argue that MIC reporting leads to misinterpretations of AST by a provider’s tendency to pick the lowest number, and thus prefer categorical reporting for simplification. Provider education and selective or cascade reporting are two strategies that may help clinicians select the optimal agent, but also have inherent limitations. All this taken into consideration, we feel the benefit of antimicrobial optimization supports reporting the MIC.

Applying Breakpoints at the Bedside

Breakpoints and MIC data are utilized at the bedside to avoid utilization of suboptimal agents due to issues with the drug and bug. For example, underlying resistance patterns may not be overtly apparent to clinicians. Depending on an institution’s reporting criteria, an E. coli isolate may be reported as susceptible to piperacillin/tazobactam despite being identified as an ESBL producer. In this instance, piperacillin/tazobactam may be suboptimal for certain infections based on available outcomes data, with some supposition that reported MIC values may be responsible for improper drug selection and treatment failures.4 Therefore, clinicians may favor alternative agents for infections despite documented susceptibility. This example also illustrates that certain resistance patterns (e.g., ceftriaxone resistance as surrogate for possible ESBL production) may help alert clinicians to underlying resistance, steering them clear of potential suboptimal agents reported as susceptible.

ID pharmacists often utilize MIC values to determine the optimal agent and dose for a given patient based on a working knowledge of breakpoint relationships and PK/PD. This may be taken a step further in the setting of an infection with a multidrug resistant organism which exhibit MICs in the intermediate or susceptible dose-dependent range. For certain drug-bug combinations with elevated MICs, it may be possible for the ID pharmacist to devise an optimized and personalized therapeutic regimen that will likely overcome the elevated MIC, either as monotherapy or utilizing potential synergistic combinations of different antibiotics.

The application of dose-dependent breakpoints, as seen with daptomycin for Enterococcus faecium, help illustrate the importance of understanding and applying dosing concepts in the context of established clinical breakpoints.1 This allows the clinician to not only optimize dosing, but also understand that an MIC at or near the breakpoint may warrant combination therapy in the setting of a serious infection. A lack of understanding of the relationship between breakpoints and dosing could potentially lead to an inadvertent assumption that lower FDA-approved doses would be reasonable despite the breakpoint being based on higher dosing strategies. These are just a small sampling of the clinical applications utilized on a daily basis to optimize a particular patient’s antimicrobial therapy.

A more global application of breakpoints occurs within antimicrobial stewardship programs. Antimicrobial stewardship focuses on the appropriate and optimal use of antimicrobials. To aid in this goal, AST reporting is key to developing clinical guidance on treating a variety of infections. Most notably, the optimization of empiric antimicrobial choices is aided by local or institutional susceptibility patterns. The cumulative antibiogram is pivotal to establishing optimal empiric antimicrobial choices across the spectrum of infectious diseases.5 Importantly, breakpoints are revised when new or additional clinical information becomes available, further illustrating the need to understand how a given institution implements and applies them. There is always the potential that from one year to the next susceptibilities may drop substantially due to implementation of new breakpoints, not necessarily an increased incidence of pathogens with a certain resistance pattern.6 In addition, the ability to adopt CLSI breakpoints often lags well beyond the release of the updated breakpoints, necessitating clinicians to be cognizant of changes and carefully assess categorical reporting as it may be discordant with the most up-to-date breakpoints.1,6 This topic will be the focus of our next blog.

More specific application of AST data can be made for either site (e.g., blood, urine, etc.), infection type (e.g., community-acquired pneumonia), or organism-specific MIC distributions. A recent trend in national guideline recommendations centers on ascertaining local susceptibility and pathogen distribution to provide optimal antimicrobial choices for certain infections (e.g., pneumonia).7 In addition, focused review of individual pathogen MIC distributions provide another level of detail that helps better define upfront institutional antimicrobial dosing strategies to optimize PK/PD parameters. In conclusion, the reporting of AST data remains a cornerstone for the effective management of infections. Information collected by laboratory technicians provides invaluable information to clinicians. This information helps optimize the selection and dosing of antimicrobial agents at the bedside, which in turn improves the clinical management of patients and infection-related outcomes.

Image 1: Clinical scenario outlining the application of MIC and breakpoint data at the bedside.

References

  1. Clinical and Laboratory Standards Institute. Development of In Vitro Susceptibility Testing Criteria and Quality Control Parameters, 5th Edition (M23).
  2. Giuliano C, Patel CR, and Kale-Pradhan PB. A Guide to Bacterial Culture Identification and Results Interpretation. P&T. 2019 April 44(4):192-200.
  3. Mouton JW, Muller AE, Canton R, et al. MIC-based dose adjustment: facts and fables. J Antimicrob Chemother 2018; 73:567-68.
  4. Pogue JM and Heil EL. Laces out Dan! The role of tazobactam based combinations for invasive ESBL infections in a post-MERINO world. E Expert Opin Pharmacother.2019 Dec;20(17):2053-57.
  5. Clinical and Laboratory Standards Institute. Analysis and Presentation of Cumulative Antimicrobial Susceptibility Test Data; Approved Guideline, 4th Edition (M39-A4).
  6. Humphries RM, Abbott A, Hindler JA. Understanding and Addressing CLSI Breakpoint Revisions: a Primer for Clinical Laboratories. J Clin Microbiol. 2019 May 24;57(6):e00203-19.
  7. Khalil AC, Metersky ML, Klompas M, et al. Management of Adults with Hospital-acquired and Ventilator-associated Pneumonia: 2016 Clinical Practice Guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis. 2016 Sep 1;63(5):e61-e111.

-James Sanders, PharmD, PhD, is an infectious diseases pharmacy specialist and assistant professor at the University of Texas Southwestern Medical Center in Dallas, Texas. His academic and research interests are focused on multi-drug resistant Gram-negative bacteria, surgical site infections and HIV pharmacotherapy.

-Marguerite Monogue, PharmD is an infectious diseases pharmacy specialist and assistant professor at the University of Texas Southwestern Medical Center in Dallas, Texas. She is interested in antimicrobial pharmacokinetics/pharmacodynamics and multi-drug resistant Gram-negative bacteria.

-Andrew Clark, PhD, D(ABMM) is an Assistant Professor at UT Southwestern Medical Center in the Department of Pathology, and Associate Director of the Clements University Hospital microbiology laboratory. He completed a CPEP-accredited postdoctoral fellowship in Medical and Public Health Microbiology at National Institutes of Health, and is interested in antimicrobial susceptibility and anaerobe pathophysiology.

AI vs. Crystallography: Predicting Pathogenic Variants?

Some very exciting news was recently announced about Artificial Intelligence impacting protein structure prediction. But like many of us, you probably thought, “Oh that’s nice.” Followed by either, “But that’s unlikely to impact lab medicine” or “I have no idea how they did that.”  Today I will help turn around those last two thoughts for you!

The big news was that a U.K. company specializing in artificial intelligence, DeepMind (owned by Google-of course), won the CASP14 competition. CASP14 is the 14th edition of the biannual bake-off competition where teams use bioinformatic approaches to predict protein structures. The organizers then judge how well predictions match experimentally derived structures using a score called: GDT. This score reflects the distance of where something is vs. where it should be. Each of the ~150 protein sequences are scored on this basis and given a final percent identity score (0-100%).

Figure 1. The Z-score is just the difference of a sample’s value with respect to the population mean, divided by the standard deviation. The groups that are markedly better than the average will have larger Z-scores.

Since the competition started in 1997, the winners have scored ~50% on average. That is until 2 years ago when AlphaFold, the AI created by DeepMind, won with a top score of 55%. Their paper was published open access (Ref: https://www.nature.com/articles/s41586-019-1923-7) and used similar techniques applied by others where proteins were progressively folded by a computer until the lowest energy state is revealed.

Figure 2. Improvements in the media accuracy of predictions in the free modelling category for the best team in each CASP. Measured as best-of-5 GDT.
GDT: Global Distance Test (0-100); the percentage of amino acid residues within a threshold distance from the correct position. GDT of around 90 is considered competitive with results obtained from experimental methods.

The programs driving this folding may consider amino acid charge, size, and polarity, genetic conservation (Ref), or similarity to other protein domains. However, the innovation here was that DeepMind used artificial intelligence to examine sequence information with a convolutional neural network to identify structural constraints that are used to predict accurate protein folding.




Figure 3. Sequence of events from Dataà Deep neural network (Artificial intelligence)àPredictions à protein folding process. (Figure 2 of this reference: https://www.nature.com/articles/s41586-019-1923-7/figures/2).

This results in one of those famous algorithms you’ve heard about. However, these algorithms are more complex than a simple linear regression and it is nearly impossible to trace how exactly how different levels of importance were assigned to each variable. An important requirement for an accurate A.I. derived algorithm is that it has a large training data set. Fortunately for Deepmind, they were able to train AlphaFold using about 170,000 structures that were determined experimentally using x-ray crystallography, nuclear magnetic resonance spectroscopy, and electron microscopy.

Although we haven’t seen what was changed between AlphaFold and AlphaFold 2, we have learned that AlphaFold 2 vastly outperformed the original in CASP14 with 91% accuracy. When programs are >90% accurate they are considered to be essentially as good as experimentally derived structures. In fact, AlphaFold 2 was able to provide more information than the experiments!  One researcher found that their experimentally derived structure had a different configuration than the one predicted by AlphaFold, so they assumed the prediction by AlphaFold 2 was incorrect. After further analysis, the experimentally derived structure was found to be very similar to the structure predicted by AlphaFold 2. In another case, AlphaFold 2 predicted that an amino acid was in an infrequently found conformation, so they figured AlphaFold 2 made a mistake. After reanalyzing the experimental data, they found that that AlphaFold 2 was correct. It was even suspected that several lower-scoring structures based on NMR data may reflect lower accuracy in the experimental structure instead of a problem with the algorithm.

Figure 4. (Left) Model for the T1064 target (red) superimporsed onto the structure from DeepMind in CASP14 (blue). (Right) Black and green structures are from the runner-ups who made predictions for the same structure (correct in blue). Obtained from CASP14 webpage on Tuesday December 1st, 2020.

Will AI replace experimental crystallography? To answer this question, I turned to a colleague in my basic science lab, Lijing Su, who has been a structural biologist for many years. Like many cases of AI, this is a useful tool, but it doesn’t entirely replace her work because a lot of the structural biology research focuses on how proteins move and change as they do their job. Structural biology has moved beyond structures of single proteins and is now focused on how different proteins interact. There is still a role for crystallographers as AlphaFold cannot perform this role…yet.

All this still begs the question of a laboratorian “Who needs to know protein structure anyways?” We understand that knowing protein structures can help explain function, which has implications with drug development. However, our main role is to provide tests that diagnose disease. A major challenge in molecular pathology is to predict whether a genetic variant causes loss of protein function. Current software has poor performance (PolyPhen2 sensitivity= 45% specificity= 50%) as they mainly measure changes in chemical properties and amino acid site conservation. One potential application of AlphaFold is to examine the effect of genetic variants on protein structure. Pathogenic changes would be predicted to deform portions of the structure impairing activity or provoking degradation through the unfolded protein response.

As the current speed of the program is quite long, this could be difficult to implement immediately, but it is imaginable that this will become quicker. A straightforward way to validate this AI software would use confirmed pathogenic or benign variants from the public database ClinVar. There are over 1,000,000 entries into this database, which would provide a useful training and validation set. It is likely that change in protein structure would be a stronger mechanism of disease for certain types of proteins (ion channels for epilepsy or myosin chains for muscular disorders) and a less strong predictor of pathogenicity for other types of proteins (enzymes for metabolic disorders or signaling proteins where protein-protein interaction is important for function).

This blog entry was written with the very helpful insights and knowledge of Lijing Su, PhD.

References

  1. Senior AW et al. Improved protein structure prediction using potentials from deep learning. Nature. 2020; 577: 706–710.
  2. CASP14 website: https://predictioncenter.org/casp14/
  3. Arnold CN et al. ENU-induced phenovariance in mice: inferences from 587 mutations. BMC Res Notes. 2012; 5: 577.
  4. https://deepmind.com/blog/article/alphafold-a-solution-to-a-50-year-old-grand-challenge-in-biology

-Lijing Su is Assistant Professor in the Center for Genetics of Host Defense at the University of Southwestern Medical Center. She specializes in structural biology and helps determine multi-protein interactions to explain unknown mechanisms of genes important to immunology.

Jeff SoRelle, MD is Assistant Instructor of Pathology at the University of Texas Southwestern Medical Center in Dallas, TX working in the Next Generation Sequencing lab. His clinical research interests include understanding how lab medicine impacts transgender healthcare and improving genetic variant interpretation. Follow him on Twitter @Jeff_SoRelle.

The Laboratory Safety Challenge

In 2014 there was an internet challenge which exploded in popularity. It was the ALS Ice Bucket Challenge in which people would dump cold water on their heads and post the video on social media. The person getting the ice water dumped on them would challenge others to post a video of their dousing and they would in turn donate to the cause of finding a cure for ALS, a progressive neurodegenerative disease. The challenge became a world-wide sensation and raised $115 million for ALS research. But, like many good things, the challenge had a dark side. Many people were injured while attempting the challenge, and at least two deaths were at least indirectly associated with it.

Another challenge has come to social media lately, and this one involves a technical skill in the laboratory. It, too, has a dark side. The blood smear challenge is the latest rage for lab techs who enjoy posting videos on Facebook, Instagram, and other social media platforms. Lab techs show off their skill by making the perfect blood smear. At first it was about who could make a smear with the most perfect beautiful, feathered edge. Then the challenge evolved into people making smears while holding the top slide with one finger or even a pencil. There are those who were quite proud to show off their skill and work.

When watching videos of people in various labs performing this challenge, I cannot help but cringe. Several of these lab techs are not wearing lab coats. Many are not wearing gloves, and I have not seen any perform the challenge while using face protection or goggles. Ignoring the safety regulations about using basic personal protective equipment is apparently the norm. These people post this online without a second thought to a public display of working in the lab without PPE. It speaks volumes about the safety culture in those laboratories, and what it says is not favorable.

The next, less obvious safety issue with the videos is that they are created using cell phones or other personal electronic devices in the laboratory. People are handling devices sometimes with gloves, sometimes without, or they are setting them on lab counters which are likely contaminated. The use of cell phones and other personal electronic devices is a dangerous infection control issue, but it is unfortunately all too common. Even before this latest challenge, lab staff all over the country pose for pictures for social media posts that are taken by cell phones. Despite the fact that known and reported infections have occurred in labs from cell phones (and other items brought home from work), techs continue to use them.  

Other issues with the blood smear challenge may be less obvious. Unless these smears are being used, valuable lab supplies are being wasted. Slides and blood-dispenser cap piercing devices cost money, and many lab supplies manufacturers have run into supply shortages this year because of the pandemic. To have a lab waste money or run into shortages for the sake of this challenge might seem foolhardy to some.

Another safety issue with the challenge is the blatant act of playing around with human, potentially infectious blood to make the smears. Staff use engineering controls, work practice controls and PPE to separate people from the hazards in the laboratory. To place oneself at risk unnecessarily, especially during the COVID-19 pandemic, borders on reckless.

When the COVID-19 pandemic began affecting labs over a year ago, many laboratorians became concerned for their own personal safety. They were unsure about how they might catch this virus and what effects it might for them and their family. These were valid concerns, and some still have fears today. In conversations with lab staff over the past months I reminded them that they work with bloodborne pathogens every day, and many are as potentially dangerous (or more) than the COVID-19 virus. If Standard Precautions are used on the job, workers will be safe from infections from COVID-19 and other pathogens. The same is true today. Laboratorians may be less worried about the coronavirus, but the risk of infection in labs from this and other pathogens is as real as ever. Using engineering controls, PPE, and safe work practices is the only way to ensure lab staff can go home without bringing something dangerous to our families.

Challenges can be fun. I participated in the ALS Ice Bucket Challenge. I came out unscathed, but I was likely just lucky, not safe. The same is true for those posting pictures and videos online from inside laboratories. You might have been working that way for years and nothing has happened. Again, that is just luck, and it will run out. Make sure you and your staff are doing what is right, and what is safe. The real challenge is how to get laboratorians in all labs to work safely and follow basic safety regulations. Can your lab meet that challenge?

Dan Scungio, MT(ASCP), SLS, CQA (ASQ) has over 25 years experience as a certified medical technologist. Today he is the Laboratory Safety Officer for Sentara Healthcare, a system of seven hospitals and over 20 laboratories and draw sites in the Tidewater area of Virginia. He is also known as Dan the Lab Safety Man, a lab safety consultant, educator, and trainer.

The Pathology Value Chain and Global Health, Part 3

In the first two installments of this blog series, we looked at inbound logistics and operations in which we can conclude that competitive advantage may be challenging to achieve. Now we turn to outbound logistics or, in simplest terms, the pathology report.

No document can be more terrifying for a patient than a pending pathology report from a biopsy, as it may contain a benign diagnosis, a malignant diagnosis, or something entirely unexpected. These reports are so important that unsuspected (non-malignant) and malignant diagnoses are included as “critical values” requiring a call and documentation to the clinical team as soon as they are discovered. Pathology reports in HIC are often not immediately available to the patient (unlike other laboratory tests) because the reports are often complex, may contain confusing terms, and may use language that patients inappropriately react to without the guidance of their clinician for meaning in their care. For example, cytology reports may be highly informative to a clinician by simply stating, “No evidence of malignancy” but may be stressful to a patient without guidance because there is not a definitive answer to what a lesion was. Similarly, a colon resection that states, “Invasive adenocarcinoma confined to the mucosa” is good news to the clinician but the first two words (and the internet) may be disturbing for the patient. The important point here is that pathology reports are written for clinicians and not written for patients as an audience. To that end, pathology reports should be highly aligned with the clinical decision-making process, an approach which is naturally aided by standardize or synoptic reporting of cancers using guidelines such as those of the College of American Pathologists, the Royal Colleges of the UK and Australia, and/or the International Collaboration on Cancer Reporting (a consortium of CAP, RCUK, RCA, ASCP, and others). These templates for a given cancer are complex, not easily committed to memory, nuanced, and require a high degree of pathology knowledge to apply correctly from the gross to the final histology findings. Thus, the value in these templates is in use by a pathologist directly, making task-shifting in this area nearly impossible without the aid of tools such as whole slide imaging and artificial intelligence (which still require a pathologist to finalize the report). Like operations, we see that a “standard of care” or a “standardized approach” to reporting cancer reduces the variability or uniqueness that can be achieved with a pathology report, infringing on competitive advantage.

Outbound Logistics – This activity covers the distribution of the final product to the consumer. For the maximum value to the patient, a report should be organized to match the treatment plan, available immediately upon completion, and provide an unambiguous answer than can be acted on. Although the first two activities generate the most important information for the patient and do so with “standards of care”, this activity involves communicating the results to the clinical team members who will act on it and, therefore, can open opportunities for competitive advantage. A new diagnosis of cancer is considered a “critical value” and requires a communication with documentation to the clinical team. However, much of pathology’s role in cancer care includes work with existing cancer patients so rapid communication of any result (not just the first cancer diagnosis) can add value. For example, integration of the pathology laboratory information system into the electronic medical record creates immediate results to clinicians. Alert systems including text messages, instant messages, emails, faxes, etc. add value by informing the busy clinician that the result is there. Photographs of the tumor grossly, histologically, or the results of specials studies can be included in printed or digital reports. Pathologists can attend tumor boards or other in-person or virtual meetings to present the results and explain them if there are questions. The more information that is transmitted with clarity to clinicians, the higher value the patient will obtain. The challenge in this activity is that the payment for the laboratory services ends with the diagnostic report and appropriate coding and, thus, laboratories may have to upcharge for their services to add these features. These further communications, which we can see adds value to the patient, does not add value to the laboratory’s revenue model without upcharges. In fact, it likely costs more to have such active communications as it takes pathologists away from the higher volumes which do equate to higher revenue (as we saw in operations). Streamlining these types of communications with electronic systems is key in cost and time savings and is the basis for the laws and regulations, for example, in the USA which require electronic medical records including laboratories. However, as laws, regulations, and guidelines evolved, these electronic communications are becoming standard of care requiring the entire system to increase the costs to have them but eroding the competitive advantage of providing such concierge services. Consider the change COVID-19 has had on communications between patients, clinicians, and the laboratory where a multi-person discussion of a case with images and consensus opinions can be done in a few minutes over a video conference without anyone leaving their office. Has this crisis provided a new way to capture time (and therefore revenue) but still provide concierge services? Or has it (more likely) created a new normal that everyone has to adopt (eroding competitive advantage)?

When we turn to LMICs and observe the activities of the pathology laboratory, communication with clinical teams on the front or back end has been uncommon and traditionally not done. Oncological practices in HIC are filtering down to LMICs including tumor boards, frozen sections (i.e., rapid, in surgery diagnostics), etc. and being instituted with some frequency. These activities improve patient value and outcomes, educate the teams in both directions, and are clearly beneficial to the system. But they take time and effort away from already understaffed systems which detracts from the value of other patients ultimately. However, when we observe these systems, we often find that they lack electronic tools for running the laboratory internally which inhibits tools for reporting externally. Thus, the major needed solution now is that any histology laboratory anywhere in the world should be using an anatomic pathology laboratory information system as it creates internal and external tools for standardized reporting, communication, and management. Furthermore, it creates better opportunities to integrate synoptic (templated) reporting, interdisciplinary team activities, and standardization of requisitions (i.e., upon receipt of samples). Greatly increased value for patients in LMICs can be achieved with electronic APLIS.

Lastly, there are incredible examples of pathologists who make time in their day to meet with patients to discuss their pathology reports. These discussions can only focus on what the reports says and what the words in the report mean, as defined not in context of that patient. Such exchanges can provide patients with helpful questions to ask their clinicians and prepare them to better understand what the clinicians suggests as next steps for treatment. Clearly valuable to the patient, these exchanges are also valued by the pathologists who enjoy the face-to-face interactions with patients that humanize the process. In rare cases (possibly a for-profit situation), these services may generate revenue but under current medical billing rules there is no standard mechanism for the pathologist to be reimbursed. If we have identified this as adding value to the patient in the pathology value chain, should we not try to find ways to build these services into the care model financially? With the ubiquitous use of video conferencing in the COVID-19 era, can this task be of minimal effort to pathologists but still add value for patients?

In our last installment, we will discuss marketing & sales and service, both of which are particularly flawed and fascinating to consider.

References

  1. Porter, M. (1985). The value chain and competitive advantage, Chapter 2 in Competitive Advantage: Creating and Sustaining Superior Performance. Free Press, New York, 33-61.
  2. Histology. Wikipedia. https://en.wikipedia.org/wiki/Histology#:~:text=In%20the%2019th%20century%20histology,by%20Karl%20Meyer%20in%201819.
  3. Thorpe A et al. The healthcare diagnostics value game. KPMG International. Global Strategy Group. https://assets.kpmg/content/dam/kpmg/xx/pdf/2018/07/the-healthcare-diagnostics-value-game.pdf
  4. Digital Pathology Market CAGR, Value Chain Study, PESTEL Analysis and SWOT Study|Omnyx LLC, 3DHISTECH Ltd, Definiens AG. https://www.pharmiweb.com/press-release/2020-06-30/digital-pathology-market-cagr-value-chain-study-pestel-analysis-and-swot-study-omnyx-llc-3dhistec
  5. Friedman B. The Three Key Components of the Diagnostic Value Chain. Lab Soft News. January 2007. https://labsoftnews.typepad.com/lab_soft_news/2007/01/the_three_eleme.html
  6. XIFIN. The Evolution of Diagnostics: Climbing the Value Chain. January 2020. https://www.xifin.com/resources/blog/202001/evolution-diagnostics-climbing-value-chain
  7. Sommer R. Profiting from Diagnostic Laboratories. November 2011. Seeking alpha. https://seekingalpha.com/article/305931-profiting-from-diagnostic-laboratories#:~:text=The%20three%20year%20average%20operating,current%20operating%20margin%20of%2012.9%25.
  8. Cancer Patients Want to Pull Back the Curtain on Pathology. M Health Lab. October 10, 2019. https://labblog.uofmhealth.org/industry-dx/cancer-patients-want-to-pull-back-curtain-on-pathology
  9. Guttman EJ. Pathologists and Patients: Can we talk?. Modern Pathology. May 2003. https://www.nature.com/articles/3880797
  10. Lapedis CJ et al. The Patient-Pathologist Consultation Program: A Mixed-Methods Study of Interest and Motivations in Cancer Patients. Arch Path Lab Med. August 20, 2019. https://meridian.allenpress.com/aplm/article/144/4/490/427452/The-Patient-Pathologist-Consultation-Program-A
milner-small


-Dan Milner, MD, MSc, spent 10 years at Harvard where he taught pathology, microbiology, and infectious disease. He began working in Africa in 1997 as a medical student and has built an international reputation as an expert in cerebral malaria. In his current role as Chief Medical officer of ASCP, he leads all PEPFAR activities as well as the Partners for Cancer Diagnosis and Treatment in Africa Initiative.

ASCP Releases Two Evidence-Based Recommendations for COVID-19 Testing

COVID-19 testing can be a bit confusing. Recently, ASCP released two recommendations for COVID-19 testing to help clinicians and laboratories sort through the noise and order the right test at the right time. In addition, ASCP has a plethora of COVID-19 resources, including Town Halls, podcasts, journal articles, and more.