Jak in the Brain: An Incidental Hematologic Malignancy Diagnosed From a Metastatic Lung Adenocarcinoma by Next Generation Sequencing

In January 2020, Mr. J, a 69-year-old man presented to the emergency department (ED) after a fall at home. His medical history included a previous 40 pack-a-year smoking habit, coronary artery disease (for which he received a bare metal stent in 2014), and a left-sided video-assisted thoracoscopic surgery (VATS) lung resection in 2017 for a stage I adenocarcinoma of the lung. At the ED, he was accompanied by his wife who witnessed the fall and did not notice any preceding neurologic deficit. On exam, his GCS was 12. Imaging, including computed tomography (CT) of the head followed by magnetic resonance imaging (MRI) of the brain, demonstrated a 3 cm left-sided cranial mass in the frontotemporal junction without mass effect or bleeding. A continuous electroencephalogram (EEG) showed epileptiform discharges. The patient was started on Levetiracetam for seizure prophylaxis. CT scans of the chest, abdomen, and pelvis showed no further evidence of malignancy; the CT scan of the chest demonstrated scarring at the site of his prior surgery and emphysematous changes in the upper lung lobes. The CT scan of the abdomen showed mild splenomegaly. Neurosurgery was consulted and recommended a craniotomy and resection of the mass, which were performed. Initial pathology from the mass was consistent with a lung adenocarcinoma and the oncology service was consulted for outpatient follow-up.

Mr. J recovered well and was referred for post-operative radiation to the tumor bed. He presented to the oncology clinic for evaluation of his oligometastatic lung cancer. Further analysis of his pathology showed 90% PD-L1 positivity by IHC, a KRAS G12D mutation, a TP53 frameshift mutation, CHEK2 inactivation, CDKN1B loss, and a JAK2V617F mutation (VAF 5.66%). The primary tumor resected in 2017 during his VATS procedure had not been sent for next generation sequencing (NGS). Labs on presentation showed a leukocytosis with a neutrophilic predominance and absolute monocytosis. These had been present for at least 10 years from chart review (Figure 1) and up to 20 years according to the patient. During that time, the leukocytosis had been under observation only given the lack of associated symptoms and chronicity. Hemoglobin, hematocrit, and platelet count were within normal limits. The physical exam showed that the edge of the spleen was not palpable below the costal margin.

Following discussion at the multidisciplinary thoracic tumor board, the patient was initiated on pembrolizumab monotherapy, which he received for 14 months without issue or evidence of progression. In March 2021, progression of the disease was noted on imaging along the surgical margin of his VATS resection on CT scan and with mild increased tracer avidity in the ipsilateral lymph nodes on his positron emission tomography (PET) scan. Resection was deemed high risk due to the size of the lesion and existing emphysematous changes and radiation was felt to be of limited benefit due to previous metastatic spread and some possible hilar involvement. As the patient was being prepared to switch to carboplatin and pemetrexed therapy, peripheral blood reverse transcription-polymerase chain reaction (RT-PCR) was checked as the leukocytosis appeared to be increasing, now to consistently greater than 20,000 without evidence of infection. JAK2 V617F in the peripheral blood was positive, and quantification showed a VAF of 35% without any associated mutations in MPL or CALR. Bone marrow biopsy was pursued and showed some atypical megakaryocytes with 80% cellularity, and 1 of 3 reticulin fibrosis. Cytogenetics and FISH analysis were normal. JAK2 V617F of the biopsy sample was positive with a VAF of 28%. These findings led to the diagnosis of primary myelofibrosis (pre-fibrotic/early stage). The patient was started on carboplatin and pemetrexed, which caused an increase in his platelets following the first cycle from 403,000/mm3 to 903,000/mm3. His thrombocytosis decreased in subsequent cycles but remained high. His leukocytosis was increasingly variable with subsequent cycles of chemotherapy. Carboplatin was discontinued after four cycles, and he remains on pemetrexed with a partial response by RECIST 1.1 criteria.

Discussion

Next generation sequencing (NGS) is now one of the hallmarks for the diagnosis and management of cancer in general, and an array of treatments has been developed for solid tumors and hematologic malignancies based solely on mutational profiles. In this case, we come across the intersection of knowledge gained for the treatment of a known neoplasm (lung cancer) and we find evidence of another, though likely pre-existing, myeloproliferative neoplasm (MPN). Due to the overt sensitivity of next generation sequencing, in this case, we theorize that white blood cells (and potentially differentiated astrocytes [1]) expressing the JAKV617F mutation, were the sources of the mutation and account for the difference in VAF seen between the tumor and peripheral blood. Without the leukocytosis and mild splenomegaly, the question of whether further testing would have been beneficial to the patient is an important one. Treatment for his lung cancer also causes significant unmasking of his MPN, as since initiation of chemo/immunotherapy, his thrombocytosis and leukocytosis have significantly increased. The evidence for an MPN in this patient had been subtle, and if not for the incidental JAK2 mutation detected on resection of the cerebral tumor in this case, it is unlikely he would have received additional workup.

Clonal Hematopoiesis and Incidental Mutations

The prevalence of myeloproliferative neoplasms, such as myelodysplastic syndromes (MDS), primary myelofibrosis (MF), polycythemia vera (PV), or essential thrombocythemia (ET) in the setting of a detected JAK2 mutation is relatively high. In a study by Nielsen and colleagues of 63 individuals positive for JAK V617F, 48 (76%) were found eventually to have an MPN (2). In an article by Riedlinger and colleagues, eight individuals with solid malignancies are found to have an incidental JAK2 mutation, of whom four (50%) were diagnosed with a coexisting MPN (2 PV, 1ET, 1MF) (3). This case highlights the difficulty in classifying detected mutations as clonal hematopoiesis versus indicators of an underlying malignant process. His leukocytosis presented another reason for further hematologic evaluation, though being asymptomatic, he had never been seen by hematology and the value remained around the upper limit of normal until the diagnosis of his lung cancer (Figure 1A). Mutations noted on tumor- or blood-based sequencing or targeted mutation analysis do not always indicate an acute underlying myelodysplasia or myeloproliferative neoplasm, but many mutations like those in JAK2 do portend an increased risk for future myeloid malignancies through clonal evolution. Because they correlate with the eventual development of a hematologic malignancy to varying degrees (4, 5), detecting these mutations can affect risk stratification and planning for the treatment of solid tumors.

Due to the challenge of identifying the clinical significance of incidental mutations and their implications, a further hematologic workup for these mutations should be strongly considered. In some cases, this may need to be prior to initiating treatment for co-occurring malignancies, depending on the aggressiveness of the malignancy. In our case, we found that marrow recovery from a hematologic insult when there is an unknown hematologic condition may produce unexpected results. For our patient, Mr. J, his oligometastatic lung cancer with brain involvement took priority. However, as chemoimmunotherapy produced erythrocytosis and thrombocytosis during treatment instead of the expected cytopenias, this triggered an important follow-up. When treatment was held, leukocytosis was additionally noted between cycles. We were able to further evaluate his leukocytosis and space his chemotherapy in order to obtain a useful bone marrow biopsy. With the other factors in his history, we had a low threshold for a peripheral blood quantitative PCR, prompting a bone marrow biopsy with next generation sequencing confirming the diagnosis. In any similar presentation, a knowledge of mutations that are unique to hematologic malignancies and clonal hematopoiesis is essential, as this case could easily apply the same considerations to TET2 or DNMT3A in the NGS of a metastatic lung tumor.

In summary, with the widespread implementation of NGS and other molecular testing including cell-free DNA, we will all encounter more incidentally discovered mutations which may be clonal hematopoiesis, an MPN, or MDS. In the future, there is also certain to be more guidance on how to work up incidental mutations, if at all. As we understand more about the clinical implications of these mutations, we will be able to more accurately risk stratify individuals and recommend an earlier workup and judgements about co-existing disease. In the setting of an incidentally found JAK2 mutation in an uncommon site, this case would support further investigation into the possibility of an underlying MPN.

*Identifying details and some lab values of the individual mentioned in this case have been changed to protect the anonymity of the patient which does not, in the author’s opinion, affect the educational value or case presentation. Verbal permission was obtained from the patient for the write-up and publication of this case.

Acknowledgement: This article was edited by Maya Abdallah, MD, Adam Kittai, MD, and Kristin Koenig, MD

1. Bouzid H, Belk J, Jan M, et al. Clonal hematopoiesis is associated with reduced risk of Alzheimer’s disease. Blood. 2021:138,5.
2. Riedlinger G, Hadigol M, Khiabanian H, et al. Association of JAK2-V617F mutations detected by solid tumor sequencing with coexistent myeloproliferative neoplasms. JAMA Oncology. 2019:265-267. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6439565/.
3. Nielsen C, Bojesen S, Nordestgaard B, et al. JAK2V617F somatic mutation in the general population: myeloproliferative neoplasm development and progression rate. Haematologica. 2014:99(9):1448-1455. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4562533/.
4. Kusne Y, Xie Z, Patnaik, M. M. Clonal hematopoiesis: Molecular and clinical implications. Leuk Res. 2022:113:106787.
5. Genovese G, Kahler A, Handsaker R, et al. Clonal hematopoiesis and blood-cancer risk inferred from blood DNA sequence. New England Journal of Medicine. 2014:371:2477–2487.

Dr. Eric Vick is a member of the ASH Trainee Council. He advises and holds equity in Salomon’s House, LLC. Dr. John Steffes indicated no relevant conflicts of interest.