Leptomeningeal disease in BRIP1-mutated pancreatic adenocarcinoma
- Eunji Yim and
- Denise Leung
- Neurology, University of Michigan, Ann Arbor, Michigan, USA
- Correspondence to Dr Eunji Yim; yime@med.umich.edu
Abstract
Leptomeningeal disease is rare in pancreatic cancer and prognosis remains poor. Mutation profiles are now directing therapy to improve survival. We describe a case of leptomeningeal and brain metastasis in BRCA1 interacting protein 1, or BRIP1-mutated pancreatic adenocarcinoma with progression through several lines of chemotherapy and immunotherapy. A woman in her late 50s presented with metastatic pancreatic adenocarcinoma on liver biopsy. She achieved complete response after modified FOLFIRINOX and started a poly (ADP-ribose) polymerase (PARP) inhibitor for a BRIP1 mutation. She had recurrence at 9 months and started pembrolizumab (programmed cell death protein 1, or PD-1receptor antibody) for tumour mutational burden of 10 muts/Mb. At 10 months, she presented with lower extremity weakness and back pain. MRI revealed leptomeningeal metastases from T11 to cauda equina roots and right occipital metastasis. Cerebrospinal fluid studies revealed elevated pressure (290 mm H2O) and protein (73 mg/dL) with negative cytology. Leptomeningeal carcinomatosis was diagnosed. She began palliative radiation but died at 11 months from initial diagnosis.
Background
Leptomeningeal disease (LMD) is an aggressive metastatic process by which the pia and arachnoid mater are seeded by malignant cells originating from a solid tumour. It is most commonly described in breast, lung and melanoma cancers but remains a rare entity with an incidence of 5%–8% among solid tumours.1 LMD is especially rare in pancreatic cancer, with a reported incidence of 0.023%.2 Regardless of primary origin, the prognosis for LMD is poor. To this end, mutation profiles across the entire cancer landscape are being developed to better target therapies that can improve survival.
The BRCA1 interacting protein 1 (BRIP1) mutation has been described in breast and ovarian cancer and more recently in colon cancer.3 Only one other case has been reported in pancreatic cancer, with an incidence of 0.5% in the Cancer Genome Atlas.4 We report a case of synchronous LMD and brain metastasis in a patient with BRIP1-mutated pancreatic adenocarcinoma who rapidly progressed through several lines of chemotherapy and immunotherapy.
Case presentation
A woman in her late 50s presented with several weeks of nausea, progressive epigastric bloating and general malaise. Initial serum testing revealed mildly elevated liver enzymes (AST 156 units/L, ALT 90 units/L), elevated carcinoembryonic antigen of 13.2 ng/mL, lipase of 87 units/L and elevated carbohydrate antigen (CA) 19–9 level of 1057 units/mL (normal range 0–37 units/mL). Cancer antigen 125 was elevated at 72 units/mL. CT scans of chest, abdomen and pelvis revealed a 2.5 cm mass in the junction of the pancreatic head and body with dilatation of the pancreatic duct. There was also 2.3 cm portacaval lymph node metastasis, extensive liver metastases and several small lung nodules. Endoscopic ultrasound-guided needle core biopsy of a left hepatic lobe mass revealed moderately differentiated adenocarcinoma consistent with pancreaticobiliary or upper gastrointestinal primary, but ultimately diagnosed as metastasis from a pancreaticobiliary tumour. Immunohistochemistry staining was positive for cytokeratin 7, cytokeratin 20 and CDX2 but negative for TTF stains. There was no information on MUC1 and MUC2 expression, so intestinal subtype was not ruled out, but this was felt to be rare and unlikely, especially based on the location of the mass. The patient had an Eastern Cooperative Oncology Group (ECOG) performance status of 1 and CA 19–9 level of 2191 units/mL. She was initially considered for clinical trials, but due to elevated liver enzymes greater than five times the upper limit of normal (AST 353, ALT 23), she was not a trial candidate and instead started modified FOLFIRINOX therapy (fluorouracil, leucovorin, irinotecan, oxaliplatin without 5-FU bolus).
The patient achieved complete response after eight cycles and demonstrated an ECOG of 0 with CA 19–9 normalised to 31 units/mL. Oxaliplatin was dose-reduced for the final two cycles due to development of grade 3 distal neuropathy. After completing 10 cycles of mFOLFIRINOX, maintenance olaparib, a PARP inhibitor, was started after discovery of a BRCA1 interacting protein C-terminal helicase 1 loss (BRIP1) mutation. Eight months after initial diagnosis, she began to experience epigastric discomfort, mild fatigue and loss of appetite. CT abdomen/pelvis revealed disease progression and CA 19–9 levels had again risen to 1129 units/mL.
Nine months after initial diagnosis, pembrolizumab, a PD-1 receptor monoclonal antibody, was started due to slightly increased tumour mutational burden (TMB) of 10 muts/Mb. After two cycles of pembrolizumab, the patient reported worsening fatigue, poor oral intake and feeling a little clumsy walking around; CA 19–9 had increased to 8545 units/mL and liver markers were also elevated (AST 217, ALT 203). Her clinical status was deemed secondary to disease progression rather than immunotherapy-related hepatitis and the treatment plan was switched to FOLFIRI, omitting oxaliplatin due to grade 3 peripheral neuropathy from prior chemotherapy.
Within 1 week, at 10 months from initial diagnosis, the patient was admitted to the hospital with new lower extremity weakness, low back pain and inability to ambulate without a walker. She denied radicular pain, saddle anaesthesia, or bowel or bladder symptoms.
Investigations
MRI of the spine revealed leptomeningeal enhancement from T11 distally to the cauda equina nerve roots (see figure 1). There was also a pathological fracture at L5, and numerous osseous metastases in cervical, thoracic and lumbar vertebral bodies. MRI of the brain revealed an asymptomatic, isolated 0.3 cm, right occipital metastasis (see figure 2). Cerebrospinal fluid (CSF) studies revealed an opening pressure of 290 mm H2O (upper limit 200–250 mm H2O), protein level of 73 mg/dL (normal range 15–45 mg/dL) and three leucocytes per mm3 with 93% lymphocytic predominance; no erythrocytes were present. Cytological examination of CSF revealed an increased number of small mature-appearing lymphocytes and did not show carcinoma; however, only 2 mL of fluid was examined. Repeat staging CT abdomen/pelvis showed numerous new metastatic lesions in the liver and lesser omentum, as well as increasing number and size of retroperitoneal lymph nodes (see figure 3).
(A, B) MRI of the total spine with numerous metastatic lesions throughout the spine (C5, T1, T2, T3, T4, T7, T8, T11, T12, L1, L4, L5 and S1 vertebral bodies). (B) There is leptomeningeal enhancement involving the distal cord from the T11 level and diffuse enhancement and loss of definition of the cauda equina nerve roots.
MRI of the brain with (A, B) 0.3 cm focus of enhancement within the right occipital lobe with (C) mild associated T2/FLAIR signal. There was no obvious leptomeningeal enhancement.
CT abdomen/pelvis showing innumerable hepatic metastases, atrophic pancreas with infiltrative soft tissue and multiple retroperitoneal lymph nodes.
Differential diagnosis
This constellation of findings is not otherwise seen in inflammatory processes such as transverse myelitis or a subacute degenerative process such as vitamin B12 deficiency. The B12 level was over 900. There was no history of immunosuppression and the likelihood of opportunistic infections causing leptomeningeal enhancement seemed very low. With respect to transaminitis with clearly progressive liver metastases, this was convincing for metastatic progression of cancer rather than hepatitis secondary to pembrolizumab therapy. Brain tumour board reached the consensus that spinal imaging, suspicious CSF and precipitous neurological decline were most compelling for LMD. The need for confirmation from a repeat lumbar puncture was felt to be unnecessary.
Treatment
Intrathecal chemotherapy was not recommended due to the aggressive nature of the treatment and lack of potential benefit in the setting of systemic disease progression. Temozolomide was briefly considered due to BRIP1 loss and good penetration across the blood-brain barrier. Given the extent of disease and poor prognosis, there were many discussions between care teams and family about how best to preserve the patient’s quality of life. Ultimately, palliative spine radiation was planned for a total of 30 Gy with concomitant capecitabine. The patient had an ECOG of 1–2 at the start of radiation therapy and she temporarily showed mild improvement with escalation of steroids. After seven radiation treatments, the patient had increasing fatigue and declining oral intake. Radiation therapy was aborted at 21 Gy due to poor functional status. The patient was then transitioned to supportive care and referred to hospice.
Outcome and follow-up
As functional decline continued, bilirubin level continued to rise to 3.8. CA 19–9 level was up to 30 546 units/mL. Three weeks from diagnosis of LMD, and 11 months from initial diagnosis, the patient was admitted to hospice care and died 3 days later.
Discussion
Pancreatic cancer is the fourth-leading cause of cancer death in the USA, but leptomeningeal metastasis of pancreatic origin is much less common. There are less than 30 cases in the current literature, the earliest of which was reported in 1974.5 The American Society of Clinical Oncology has published revised recommendations stating that CA 19–9 levels can be a marker for disease recurrence in pancreatic cancer and that rising CA 19–9 levels have been correlated with shorter survival times while undergoing chemotherapy.6 In our experience, CA 19–9 levels did correlate with the trajectory of clinical response and disease progression; it demonstrated an expected rise near the end of life as her disease progressed.
While the gold standard for diagnosis of LMD remains the presence of malignant cells in CSF, only 70% reveal positive cytology on the first lumbar puncture.7 Cytology has been negative in 10%–15% of patients with unequivocal LMD and can remain negative in up to 14% even after three studies, regardless of where the CSF is sampled—lumbar, cisternal or ventricular sites.1 8 In our case, the fluid sent for analysis was only 2 mL and therefore likely to represent a false negative result in the milieu of a clinical and radiographical presentation highly suspicious for LMD.9 Opening pressure of CSF was also unequivocally elevated at 290 mm H2O even with consideration of normative values adjusted for age and BMI.10 This is consistent with that described by Olson et al, in which opening pressure can be the single abnormality on CSF examination.5 Moreover, CSF protein was mildly elevated at 73 mg/dL.
Several hypotheses have been proposed for the mechanism by which cancer spreads to the leptomeninges. When first documented in the literature, Olson et al suggested direct invasion of the choroid plexus or infiltration of the perivascular spaces of penetrating leptomeningeal vessels, or alternatively, secondary invasion of nerve roots providing access to the leptomeninges.5 Gonzalez-Vitale and Garcia-Bunuel’s analysis of the choroid plexus revealed malignant cells at the base of the plexuses but none in the plexus itself or within the lumina, which supports direct extension rather than haematogenous metastasis.11
The mechanism with the strongest support, however, appears to be that of perivascular and perineural infiltration from the axial lymphatic nodes and vessels as both likely communicate with the subarachnoid space.11 12 Kokkoris compared patients with breast and lung cancer and found that those with LMD alone had a higher incidence of vertebral or paravertebral metastasis than patients with and without parenchymal metastases, suggesting that access to the vertebrae and adjacent structures is the intermediary for centripetal propagation into the central nervous system.12 Such access may be facilitated by retrograde flow through Batson’s venous plexus or perhaps via infiltration of intrapulmonary vessels, travelling along perivascular spaces which extend through the dural and arachnoid sleeves of nerve roots into the subarachnoid and subsequently pial layers.12 Our case also supports this theory, as the patient had confirmed osseous metastases in multiple vertebral segments.
Survival for LMD in pancreatic cancer ranges widely from 0.5 months to several years but remains largely unchanged despite the use of palliative approaches, including gemcitabine, nab-paclitaxel and whole brain radiotherapy.13–16 According to 2017 clinical practice guidelines offered by the European Association of Neuro-Oncology and European Society for Medical Oncology (EANO-ESMO), the management of LMD from solid tumours carry level IV–V recommendations and should rather be derived from the primary tumour and its prior treatment.17
Germline mutations have come into focus as targeted therapies are being developed against specific genomic patterns in the oncogenic landscape. To date, these have rarely been implicated in LMD from pancreatic cancer, however. O’Connor et al reported a single germline mutation in the MUTYH gene among five patients with LMD from pancreas ductal adenocarcinoma.13 BRCA germline mutations have been reported in three cases of pancreatic cancer with brain metastasis, all three of which were simultaneously also diagnosed with leptomeningeal carcinomatosis.18 19 To our knowledge, this is the first case that describes the clinical course of a patient with leptomeningeal carcinomatosis from pancreatic cancer with a BRIP1 germline mutation. Our patient also presented with synchronous intraparenchymal brain metastasis, though the isolated occipital lesion was clinically asymptomatic.
BRIP1 codes for a protein which interacts with BRCA 1/2 during homologous recombination, but also exerts other effects independent of BRCA in the DNA repair mechanism.4 There is no available data on whether aberrant BRIP1 activity can alter the growth or spread of pancreatic cancer. We do have adjacent data in BRCA2-mutated breast and ovarian cancer with isolated LMD, which reportedly responded to olaparib therapy.20 21 In our case, however, leptomeningeal spread was discovered after maintenance PARP inhibition had been attempted. This suggests that the efficacy of PARP inhibition may not extrapolate as readily to the BRIP1-mutation, but further studies will be necessary to characterise the role of BRIP1 and whether this is a viable therapeutic target.
Brastianos et al described the safety of using pembrolizumab in patients with leptomeningeal carcinomatosis from solid tumour malignancies, 60% of whom were alive at 3 months after receiving at least one dose.22 In our case, the patient received two doses of pembrolizumab, both of which predated clinical signs of developing leptomeningeal carcinomatosis and yet her disease continued to progress. Moreover, our patient had a more favourable ECOG performance score than the patients enrolled in the phase II trial yet did not survive beyond 2 months after PD1-directed therapy. One possible explanation for the lack of response is that our patient had a borderline TMB of 10 muts/Mb, the cut-off used in the KEYNOTE-158 trial.23
This is a case of pancreatic adenocarcinoma with a unique histopathological and mutational profile that rapidly progressed into extensive LMD. The current standard of care, modified FOLFIRINOX, was ultimately ineffective against disease progression, as were targeted PD1 and PARP inhibitor therapies. The significance of a BRIP1 mutation in this clinical context has yet to be elucidated and we propose further studies be conducted in the broader investigation of germline mutations as potential targets for therapy.
Patient’s perspective
While the rest of the world’s attention in 2020 was focused on the rise of COVID-19 and its associated fears, my wife was dealing with the simultaneous tragedy of learning that she had inoperable advanced stage pancreatic cancer that had spread to her liver. It was the worst possible time to be an immunocompromised chemotherapy patient. However her response to FOLFIRINOX was very encouraging and she looked and felt back to normal within four months, when we managed to have a beautiful outdoor wedding for our daughter. We understood that chemotherapy resistance was inevitable but were hopeful that the second and third line treatments would be effective. We had no idea what was happening when she suddenly lost all her strength in her legs and began the rapid decline which was the result of the rare migration of the cancer to her CNS. One small comfort was that we didn't prolong the radiation treatments which were clearly not helping and were able to limit her time in the hospital and get her home for the holidays where she was able to spend her final days surrounded by her family. Our family is very hopeful that the case study will help in some way with the advancement of knowledge on this phenomenon and perhaps encourage research directed at this terrible disease.
Learning points
-
Leptomeningeal carcinomatosis from pancreatic cancer is rare but can present synchronously with parenchymal brain metastases.
-
This is the first documented case of leptomeningeal disease (LMD) with a BRIP1 germline mutation.
-
Disease progressed through modified FOLFIRINOX, targeted PD1 inhibitor and PARP inhibitor therapies before diagnosis of LMD.
-
Further studies are necessary to better characterise whether germline mutations are viable therapeutic targets in LMD.
Ethics statements
Patient consent for publication
Acknowledgments
We would like to thank Dr John Kim for his assistance in preparing the images, Dr Henry Appelman for reviewing the pathology and diagnosis and Dr Moh’d Khushman for his expert opinion regarding differentiation between upper gastrointestinal and pancreatic cancers.
Footnotes
-
Twitter @DeniseLeungMD
-
Contributors EY: substantial contributions to the conception of the work, drafting the work. DL: substantial contributions to the conception of the work, revising it critically for important intellectual content. Both authors (EY, DL) gave final approval of the version published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Case reports provide a valuable learning resource for the scientific community and can indicate areas of interest for future research. They should not be used in isolation to guide treatment choices or public health policy.
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2022. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer