At its core, the Munich Leukemia Laboratory (MLL) is like any other diagnostic lab.
“We receive samples and send reports,” one of its three cofounders, Torsten Haferlach, MD, PhD, told ASH Clinical News. “It is basic stuff from that perspective.”
The day-to-day operations of the MLL, however, are anything but basic. Through automation, cloud computing, artificial intelligence, and state-of-the-art molecular diagnostics, the lab can process patient genome data in 3 hours – a huge step forward from the 13 years it took to map the first human genome.
The MLL’s mission is to improve the care of patients with leukemia and lymphoma through cutting-edge diagnostics. To do this, the lab uses an interdisciplinary approach combining six disciplines – cytomorphology, immunophenotyping, chromosome analysis, fluorescence in situ hybridization (FISH), molecular genetics, and bioinformatics – resulting in one comprehensive integrated laboratory report.
“In hematology, we are on the way to defining malignant diagnoses in much more detail, putting us 5 to 10 years ahead of solid tumor diagnostic tools,” Dr. Haferlach said. “We are working toward specific breakdowns of morphology and genetics to better classify leukemias and lymphomas to aid in classification of disease and improving targeted treatments.”
And their work has only just begun. ASH Clinical News spoke with MLL’s cofounders about their innovative approach to leukemia and lymphoma diagnostics and their vision for the future.
The MLL opened in August 2005 to address advancing leukemia diagnostics. Dr. Haferlach, along with cofounders Claudia Haferlach, MD, and Wolfgang Kern, MD, recently celebrated the lab’s 15th anniversary – although, due to COVID-19, not with the party its founders would have liked.
“We realized [15 years ago] that the need for diagnostic tools and combination methods was increasing quickly and there was no space for us within a university infrastructure to expand our lab and hire new personnel,” Dr. T. Haferlach said of the motivation behind creating the MLL. “We moved out of the quiet, comfortable zone of a university hospital into a private setting that combined a private lab with a hematology practice.”
In the 15 years since its opening, the MLL has processed more than 750,000 samples from settings throughout the world. The number of employees has grown from just 29 to more than 200. Dr. T. Haferlach is Head of the Cytomorphology Department, Dr. C. Haferlach is Head of the Cytogenetics and FISH Department, and Dr. Kern is Head of the Immunophenotyping Department. Manja Meggendorfer, PhD, MBA, is Head of Molecular Genetics and Research and Development at MLL, which represents the fastest-growing segment of the lab’s work.
The unique advantage of the MLL is its interdisciplinary approach. Rather than sending samples to five different laboratories that perform five types of testing spread across five cities, clinicians can take advantage of several types of testing housed in a single lab, Dr. T. Haferlach explained.
The MLL also sets itself apart by hiring not only lab scientists, but also hematologists.
“Each sample we receive is reviewed by one of our hematologists to see if the ordering physician is asking the right questions,” Dr. Kern said. “This means that we occasionally change the workflow of the sample and take the opportunity to work with the [test-ordering] hematologist and say, ‘You can ask that question, but we suggest you ask this question instead.’”
The upfront collaboration, called “order control,” avoids unnecessary diagnostics and wasted time and money, making the whole diagnostics process quicker and cheaper.
When a sample arrives at the MLL, it enters a world of extensive automation and digitization. It is added in the MLL’s database, then labeled with a barcode unique to each patient’s blood or bone marrow, and finally forwarded for cell lysis for amplification.
Next, robots divide the samples into equal portions and place them in barcode-labeled vials. These vials are automatically frozen in large refrigerator units. The next morning, the robot automatically retrieves these vials for further analyses in a 96-well plate format.
Other machines prepare the DNA, and another set of devices prepare the samples for sequencing machines. The barcode assigned to each sample can be read at all workstations and devices throughout the lab. The lab’s database is automatically updated as the sample is transferred between staff and equipment, allowing anyone querying findings about the sample to receive the latest information. “Each sample can be tracked every second through every machine and technician,” Dr. C. Haferlach explained.
Once sequencing is completed, and as each new piece of information is entered into the lab’s database, an automated system defines next steps for processing, measurement, assessment, and diagnosis. Samples sent to the MLL may undergo one or more of the diagnostic methods available at the lab:
- Cytomorphology: The basic diagnostic technique for evaluation of hematologic diseases and staging of lymphoma, cytomorphology involves assessment of staining in blood and bone marrow smears. Technicians and hematologists can assess cellularity and distribution patterns.
- Chromosome analysis: A standard test for the diagnosis of hematology neoplasia, chromosome analysis – or karyotyping – can help identify and characterize chromosome aberrations of prognostic relevance.
- Immunophenotyping: Peripheral blood or bone marrow aspirate are used to characterize and quantify benign and malignant cell populations via flow cytometry. This process uses antibodies to identify cells based on the types of antigens or markers expressed on their surface.
- FISH: More rapid than chromosome analysis by karyotyping, this diagnostic method uses fluorescent probes to identify or exclude a specific known or suspected cytogenetic aberration in metaphases or interphase nuclei.
- Molecular genetics: This method uses DNA or RNA from peripheral blood or bone marrow to identify gene mutations and fusions and determine measurable residual disease.
- Bioinformatics: This method incorporates computer science and biology for the organization, processing, and analysis of next-generation sequencing data.
“In general, we can distinguish disease subtypes through phenotypic methods like cytomorphology, immunophenotyping, and genetic techniques. In the future, molecular techniques will be the most important methods in helping us capture all this information,” said Dr. C. Haferlach. “Today, genetic diagnostics relies on cytogenetics and mutation analysis. Soon, we will be doing more whole-genome and whole-transcriptome sequencing.”
Whole-genome sequencing can identify more structural variants and copy number changes than are evaluable through chromosome analysis, and it can also be used for mutation detection and analysis, she explained.
“There could be one assay that does everything,” she predicted. “If we also include gene-expression profiling in our diagnostics, we can even catch features that are now determined by immunophenotyping and cytomorphology. This is our vision for the future.”
Research and Collaboration
To make this vision a reality, the MLL is heavily involved in research, conducting a large amount of in-house research with no financial assistance from grants or public entities. For example, the lab has sequenced thousands of genomes and transcriptomes from patient samples in its own database and freezers. The lab estimates that, for approximately 40% of sample submissions, the patient has also consented to let the lab preserve their biological material and use it as needed for research purposes.
More powerful computing and unlimited data capacity through cloud storage has helped the lab accelerate its research: Technicians can run up to 18 patients per whole-genome run, and the lab runs up to 150 genomes per week.
“We have more than 3 petabytes of data from the genomes and transcriptomes taken from more than 5,000 cases that are completely characterized,” Dr. T. Haferlach said. “Now, we can look back to see how the genome sequencing can add more information or support the information we gain from the routine techniques we use.” The MLL is using cloud computing and artificial intelligence tools to accomplish these research goals.
The MLL’s research findings are published in international, peer-reviewed journals and presented at medical conferences and scientific events. Its cofounders also acknowledge the importance of moving research advances into patient care.
“We publish a lot of our research results on our website and even have teams screening medical literature for important new advances that we then put online so that our colleagues in hematology have necessary information with respect to diagnostics and treatment,” said Dr. C. Haferlach. “We also include this type of information in our reports. We want to integrate this knowledge so everyone can put it to use.”
Recently, the MLL introduced an “Integrated Findings” service, which provides physicians and patients with an overview of the individual findings from cytomorphology, flow cytometry, cytogenetics, and molecular genetics needed for disease diagnosis and prognosis. The report also includes designation of the diagnosis according to World Health Organization classification and various prognostic scoring systems. So far, Integrated Findings are being prepared for patients with acute myeloid leukemia, and the lab is working on versions for patients with myelodysplastic syndromes and other diseases.
The MLL collaborates with researchers from several partner organizations, including hospitals throughout Europe, Japan, and the U.S. “We also invite people from our partner organizations to come here and learn our techniques,” Dr. T. Haferlach said. “We share our data and our samples. We do not throw anything away. We store more than 2 million vials of samples in our freezer. If there is something very rare that someone wants to study, we have it.”
Balancing all these diagnostic and reporting processes means the MLL needs all the help from technology that it can get. When the lab began operations, the founders worked to automate nearly every wet lab process possible. Now they are automating analysis, too.
“We completely automated molecular genetics,” Dr. C. Haferlach said. “With our staff bioinformaticians, we reviewed all the available databases, plus our in-house data, to identify the most important aspects for diagnosis and interpretation of the mutations.”
In cytogenetics, for example, the MLL developed an artificial intelligence (AI)–based algorithm to perform karyotyping, instead of having a morphologist review metaphase spreads manually. AI also has been incorporated into pilot trials in immunophenotyping and cytomorphology, where classification of the cells is done based on AI.
However, according to Dr. T. Haferlach, the most important technology used to streamline workflow at the MLL is its proprietary laboratory information and management system (LIMS). The system is constantly updating itself. It has four rollouts per day, meaning when a technician comes to work on a certain day, he or she will not be using the LIMS version from the day before.
He noted that about 50% of the machines at the MLL are organized by the LIMS system. Most devices in the lab are interconnected via sensors, software, and other technologies, saving time and minimizing the risk for human error.
The MLL team currently includes 10 hematologists, 18 bioinformaticians, 120 technicians, and 45 biologists, molecular biologists, and cytogeneticists, along with several nurses in its hematology practice. Combined, the personnel have more than 1,450 years of knowledge and experience in the field, but, if anyone leaves the lab, their knowledge leaves with them. “That means we need to try to get all of this information into workflows driven by AI, cloud computing, or other tools so it is not lost,” Dr. T. Haferlach said.
Lab personnel have worked to make the state-of-the-art diagnostics used at the MLL reproducible through the development of standard operating procedures (SOPs).
“We are accredited according to the highest standards in Europe: ISO 15189 and 17025,” Dr. Kern said. These standards were developed by the International Organisation for Standardization’s Technical Committee and signify that a medical lab meets requirements for quality and competence and generates valid results. The MLL earned these accreditations in 2009 and has maintained them ever since.
“For all the steps where we apply diagnostics, we have validated procedures and documented SOPs,” Dr. Kern said, adding that the staff performs quality assessment on a daily basis to ensure the machines are operating adequately. “We also are participating in all proficiency testing programs to get external quality controls and, in fact, all of us here are heading proficiency programs in the context of our respective programs. We are quite heavily involved in quality management.”
The MLL currently has 620 SOPs, totaling more than 4,500 pages of documents – all digital, of course.
Ready for the Future
The MLL has accomplished much in its 15-year history, but it wants to continue to shape the future of hematologic diagnostics and therapy. This will require a step-by-step process to continuously improve diagnostics and to bring those improved diagnostics to patients, according to Dr. C. Haferlach.
“We are professors and researchers,” Dr. T. Haferlach echoed. “We want to always be on the cutting edge of what is next.”
That desire has pushed the founders to fully embrace and incorporate technology into their lab.
“This [work] puts us in a position to increase significantly our case volume load while maintaining the quality at which we operate – without the need to increase personnel,” Dr. Kern said. “We are prepared for the future and for much more work to serve doctors and their patients.” —By Leah Lawrence