Persistent hypoxemia in a child with medulloblastoma following a third autologous stem cell transplant

Persistent hypoxemia in children with an underlying diagnosis of cancer presents a clinical challenge for physicians. Hypoxemia may arise directly from the malignancy, the subsequent treatment the child is receiving, or from complications associated with curative intent. The complex interplay of these factors necessitates a multidisciplinary approach to diagnosis and management, ensuring that each potential etiology is sufficiently examined and addressed. This report seeks to explore the case of persistent hypoxemia in a child with medulloblastoma following bone marrow transplant, emphasizing the importance of a systematic and comprehensive evaluation to guide both the diagnosis and the treatment. By better understanding the potential etiologies of hypoxemia, clinicians can improve patient care and meet the unique needs of this specific population. A 3-year-old boy with disseminated Group 4 medulloblastoma status postmidline craniotomy, emergent, and abbreviated course of spinal radiation followed by induction chemotherapy (vincristine, methotrexate, etoposide, cyclophosphamide, and cisplatin) developed new-onset hypoxemia on Day +4 following his third and final autologous hematopoietic stem cell transplant (HSCT) conditioned with carboplatin and thiotepa. A chest X-ray (CXR) revealed bilateral hazy pulmonary opacities and a chest computed tomography (CT) revealed diffuse bilateral peripheral ground glass opacities and central bronchial wall thickening. The heart size was noted to be normal. He underwent a bronchoscopy with lavage the following day which revealed mild visible evidence of inflammatory airway disease consistent with diffuse bronchitis, but absence of inflammation on bronchoalveolar lavage (BAL). A central venous catheter blood culture and BAL revealed pansensitive Serratia marcescens and he completed a course of cefepime for 10 days for bacteremia and possible pneumonia. Despite the 10-day cefepime course for bacteremia and possible pneumonia, he remained intermittently hypoxemic with an escalation in oxygen requirement (2 L nasal cannula [NC] 100% FiO2) on Day +17 with new-onset tachypnea and dry cough. A repeat CXR noted improved hazy bilateral opacities but new cardiac enlargement. An echocardiogram (ECHO) revealed a severely dilated right ventricle with moderately decreased systolic function with markedly increased estimated right ventricular (RV) systolic pressure (67 mmHg). A CT angiogram was negative for pulmonary embolism (PE) but did demonstrate new circumferential pericardial effusion. Focal dilation of left lower lobe pulmonary arteries with ground glass opacities was suspicious for regional venous obstruction though infection required exclusion (Figure 1). He continued to receive scheduled diuretic therapy to maintain a slight net negative balance with good effect. Cardiology was consulted and recommended maintaining oxygen saturation levels above 92% to prevent worsening pulmonary resistance from his previous pneumonia and to promote pulmonary vasodilator effect. A baseline N-terminal pro-brain natriuretic peptide was obtained which was elevated at 13,775 (normal 23–289 pg/mL). A pulmonary hypertension specialist was consulted, and the prevailing concern was for pulmonary arterial hypertension due to an acute pulmonary process. Despite these measures, the child continued to have an oxygen need (maximum 10 L via high-flow NC 100% FiO2) and underwent a repeat bronchoscopy with BAL on Day +22, which revealed essentially normal findings with no visible evidence of inflammatory lower airway disease or hemorrhage. Infectious disease studies were negative including bacterial culture, legionella polymerase chain reaction (PCR), aspergillus antigen, adenovirus PCR, beta-d-glucan, pneumocystis PCR, fungal culture, acid fast smear for mycobacteria, nocardia stain, and respiratory viral pathogen panel. A venous blood gas revealed normal CO2 (41–51 mmHg) and HCO3 (18-26 mmol/L). During routine anesthesia for the bronchoscopy, the child experienced sustained hypotension despite normal oxygen saturation of 94% requiring vasopressin suggesting a pulmonary hypertensive crisis. A repeat ECHO the following day continued to demonstrate a severely dilated right ventricle with severely decreased systolic function with further increasing estimated RV systolic pressure (79 mmHg). He was subsequently initiated on sildenafil 5 mg (0.33 mg/kg) three times a day and escalated to 10 mg (0.67 mg/kg) three times per day as tolerated based on recommendations from cardiology. Cardiac catheterization was considered but deferred at that time due to initiation of sildenafil. Challenge point: A 3-year-old boy presents with increasing oxygen requirements on Day +17 after third and final autologous stem cell transplant. CT demonstrates pericardial effusion. ECHO reveals severely dilated right ventricle with decreased systolic function. Infectious disease workup is negative. Learner reflection: What is your leading and differential diagnosis? What do you recommend next and why? Listen to podcast 1 to hear the consortium's decision-making progress. In conjunction with a pulmonary hypertension specialist and pulmonologist, the differential was broadened to include pulmonary veno-occlusive disease (PVOD), PE (however CT angiogram negative although could not exclude thrombotic microangiopathy) and pneumonia (however repeat negative BAL culture and absence of inflammation). Due to the child's persistent hypoxemia and tendency towards pulmonary edema with diuretic responsiveness, his oxygen needs were ultimately felt to be characteristic of PVOD. Thus, the child began defibrotide (6.25 mg/kg q6h) on Day +29 posttransplant due to the possibility of PVOD and pulmonary thrombi contributing to his oxygen needs. Anticoagulation was deferred due to risk of alveolar hemorrhage. Though considered, a diagnostic lung biopsy was likewise deferred given anesthesia risk. A repeat ECHO 5 days later revealed a moderate-severely dilated right ventricle with slightly improved but severely decreased systolic function with estimated RV systolic pressure (59 mmHg). Catheterization was discussed at length with an anesthesiologist, an intensivist, a cardiologist, and a pulmonary hypertension expert. Ultimately, and because he had begun empiric treatment for PVOD, the decision was made to forgo cardiac catheterization as the child demonstrated significant right ventricle dysfunction raising the concern for a postanesthesia pulmonary hypertensive crisis and hemodynamic collapse requiring intubation, which would have been detrimental for this child. For him, the risks outweighed the benefits. As he was not a candidate for a cardiac catheterization to accurately measure his RV systolic pressure, he was transferred to the pediatric intensive care unit for right ventricle optimization with milrinone (0.25 mcg/kg/min) and epinephrine (0.01–0.02 mcg/kg/min) until repeat ECHO and CT showed improvement in systolic function. A 7-day course of methylprednisolone (2 mg/kg) was initiated a week later to improve lung function. During this time, the child was gradually able to tolerate oxygen wean to 0.25–0.50 L oxygen supplementation and continued to receive diuretic therapy. A repeat ECHO a week later revealed RV size and function improvement, decrease in tricuspid regurgitation, and decreased pericardial effusion (Figure 2). A repeat chest CT demonstrated decreased patchy ground glass opacities in both lungs and resolved pericardial effusion. The child completed a 14-day course of defibrotide. There were no adverse events such as bleeding during this time. The diagnosis of PVOD was supported by resolution of cough and alveolar opacities with steroids and defibrotide, resolution of pulmonary edema, and marked improvement in RV size and function. The child was discharged home with 0.25–0.5 L oxygen supplementation, to wean as tolerated with a goal of 94% oxygen saturation. The child is currently 12 months from completion of therapy, doing well and remains in room air without oxygen support. ECHO performed at 8 months after completion of therapy revealed normal heart size and function with no evidence of pulmonary hypertension. Challenge point: A 3-year-old boy with presumed PVOD is treated with defibrotide and steroids along with milrinone and epinephrine. The child experienced significant improvement in oxygen saturation. Repeat ECHO and CT studies show improvement as well. Learner reflection: How do these treatments address the underlying pathophysiology of PVOD? What are the advantages and disadvantages of this form of treatment? Listen to podcast 2 to hear the consortium's decision-making process. This child's unclear pulmonary presentation and subsequent directed management of suspected PVOD with improvement highlights the importance of maintaining a high suspicion for PVOD in children who undergo multiple autologous or other high risk HSCTs. PVOD is a progressive occlusion of the pulmonary venous system characterized by occlusion of the small pulmonary veins by fibrous intimal thickening in addition to patchy capillary proliferation.1 PVOD can progress to increased pulmonary vascular resistance leading to pulmonary vascular congestion, progressive dyspnea, and RV heart failure.2 Clinical presentation of PVOD is characterized by fatigue and shortness of breath.3 Other findings include decreased exercise tolerance, chest pain with exertion, dizziness, and cough.3 Additionally, the difficulty of diagnosis is exacerbated by the fact that PVOD often presents with vague symptoms, nonspecific findings on chest radiographs, and atypical cardiopulmonary hemodynamics.2 Definite diagnosis is obtained via lung biopsy; however, this carries substantial risks for children with pulmonary hypertension and after individual assessment of risks and benefits, biopsy is often deferred.3 In pediatrics, the actual incidence of PVOD is unknown due to the complicated nature of diagnosis—many are misdiagnosed as idiopathic pulmonary hypertension.3 The annual incidence is estimated to be 0.1–0.2 cases per million. A broad literature search identified 68 cases of pediatric PVOD.3 In the post-HSCT setting, less than 20 pediatric PVOD cases were identified.4 Diagnosis of PVOD is particularly challenging as many findings are nonspecific.5 However, findings on radiographic imaging and on clinical presentation should prompt consideration for PVOD post-HSCT.6 Noninvasive diagnostics include clinical presentation, thin-slice chest CT with contrast, and Doppler ECHO. Findings on CT suggestive of PVOD include mediastinal lymph node involvement, centrilobular ground glass opacities, smooth thickening of the interlobular septa, and pleural effusion.3 Although the child did not have mediastinal lymph node involvement or interlobular septa thickening, he did have an original CT that revealed diffuse bilateral peripheral ground glass opacities and central bronchial wall thickening which correlates with these PVOD findings. Furthermore, pulmonary arteries were regionally dilated in areas of ground glass opacities suggestive of patchy pulmonary edema due to pulmonary venous obstruction. In addition, a subsequent CT demonstrated a new circumferential pericardial effusion which is also often seen in children with PVOD; although, the mechanism is unclear.3 Findings on Doppler ECHO suggestive of PVOD include elevated tricuspid regurgitation velocity with or without right sided chamber dysfunction.1 This child developed acute pulmonary hypertensive changes with severely dilated right atrium and ventricle with severe RV dysfunction secondary to near-systemic RV systolic pressures, which is similar to other PVOD case report findings.7 The fact that the child had so much right heart dysfunction in the absence of RV hypertrophy is likely indicative of the acuity of onset. The combined findings on CT and ECHO are suggestive of a PVOD diagnosis for the child. Diagnosis of PVOD was aided by the rule-out criteria of negative BAL cultures for pneumonia and negative CT angiogram for PE. Response to treatment is also supportive. Challenge point: The diagnosis of PVOD is often confirmed with biopsy. In this case, biopsy was not possible. Diagnosis was made utilizing imaging such as CT and ECHO. Findings here were consistent with PVOD. Rule-out criteria also contributed to the diagnosis of PVOD. Learner reflection: How do these noninvasive diagnostic tools aid in identifying PVOD? Which findings were crucial in the diagnosis of PVOD? Why are rule-out criteria also important in this case? Listen to podcast 3 to hear the consortium's decision-making process. One of the greatest risk factors for developing PVOD is endothelial injury from cytotoxic chemotherapy and irradiation.6 Recent research is also suggestive of infection playing a role in development of PVOD.8 Before the child's episode of hypoxemia leading to the PVOD diagnosis and treatment with defibrotide, the child had received spinal radiation along with chemotherapy which utilized the drugs carboplatin and thiotepa for his disseminated medulloblastoma. The use of cytotoxic drugs in this case report is consistent with other reported cases of PVOD post-HSCT.2, 8 It is possible that the spinal radiation and cytotoxic chemotherapy treatment caused extensive endothelial injury which increased the child's risk factors for developing PVOD. Additionally, infection has been shown to cause reduction in tissue plasminogen activator.9 Subsequently, this reduction may cause thrombus formation leading to intimal fibrosis of the small veins and venules. The child recovered from a Serratia marcescens infection before the decline of his hypoxemia and it can be reasonably hypothesized that this infection further predisposed the child to acquiring PVOD. The onset of PVOD after transplant typically occurs within weeks to months, with a few case reports of PVOD manifesting within days.2 This timeline is similar to the child's presentation with new oxygen requirements on Day +17 posttransplant. Challenge point: Endothelial injury contributes to the underlying pathophysiology of PVOD. Injury to the pulmonary endothelium can come from different sources such as cytotoxic injury or infection. In this case, the child underwent chemotherapy, radiation, and had a previous infection. Learner reflection: How might these risk factors contribute to endothelial injury and subsequently increase the likelihood of developing PVOD? Listen to podcast 4 to hear the consortium's decision-making process. Other noninvasive diagnostics for PVOD include resting hypoxemia, severe desaturation upon exercise, and occult alveolar hemorrhage on BAL.1 This child did not have alveolar hemorrhage, as PVOD was suspected early and rapidly treated with empiric high dose steroids. Pulmonary hypertension is often treated with pulmonary vasodilators; however, in the setting of pulmonary venous obstruction, the use of vasodilators can increase congestion in the pulmonary capillary bed, resulting in progressive pulmonary edema. Therefore, vasodilators are often contraindicated in treatment of PVOD.8 The child did receive sildenafil, but fortunately did not worsen as his PVOD was likely patchy and regional. One of the main challenges of this case was to maintain the child's oxygen at an adequate level. The child had resting hypoxemia and thus required supplemental oxygen to maintain adequate saturations. The hypoxemia combined with the tendency of the child towards pulmonary edema with diuretic responsiveness contributed to the PVOD diagnosis. Lung biopsy is often used as confirmation for a diagnosis of PVOD.5 Pathology of lung tissue in a person with PVOD characteristically contains primary venous-based lesions without plexiform lesions. Other findings include secondary arteriolar medial hypertrophy, intimal proliferation, and alveolar hemorrhage.5 Nevertheless, biopsy was contraindicated in this case given the child's oxygen needs and anesthesia risk. Placing the child under anesthesia might have precipitated a hypertensive crisis and subsequent hemodynamic collapse with concern for prolonged intubation. This decision was made to empirically treat the patient instead, with input from a pulmonary hypertension specialist and pediatric pulmonologist. Currently, there is still no clear treatment for PVOD. Defibrotide, an oligonucleotide with antithrombotic, fibrinolytic, and endothelial reparative properties has emerged as a treatment for hepatic veno-occlusive disease. Clinical studies utilizing defibrotide for people with hepatic veno-occlusive disease have been shown to be successful. A study of 45 pediatric cases with hepatic VOD demonstrated a 76% complete resolution of VOD with defibrotide treatment.10 Other studies show similar favorable results of defibrotide treatment for hepatic VOD.10 Moreover, a phase III clinical trial reported a 23% increase in survival rates at Day +100 post-HSCT in people with hepatic VOD.11 Defibrotide is a drug that selectively increases prostaglandin I2 and E2 levels in addition to increasing tissue plasminogen activator. These increased prostaglandins and tissue plasminogen activators work by increasing the fibrinolytic capacity of endothelial cells while also reducing their procoagulant activity.12 Pulmonary alveolar hemorrhage was the most common adverse effect with a fatal outcome.11 The child, however, did not develop hemorrhage postdefibrotide treatment. Due to the pathophysiologic similarities between the hepatic and pulmonary forms of veno-occlusive disease, it may be hypothesized that defibrotide would have a similar benefit in treating PVOD. Nonetheless, medical literature does not yet contain any cases reporting the use of defibrotide to successfully treat PVOD.13 Challenge point: A 3-year-old boy presented with signs and symptoms characteristic of PVOD. After treatment with defibrotide and steroids, the child completely recovered. The child is now at home without any oxygen requirements. Learner reflection: What are the potential implications of the positive outcomes of this case for future research and treatment approaches for PVOD? Listen to podcast 5 to hear the consortium's decision-making process. This particular case report details the use of defibrotide in a child with PVOD post-HSCT. This child was successfully treated with defibrotide with resolution of PVOD. The child is doing well in the outpatient setting without oxygen support and without any major respiratory concerns. Although the child had findings not consistent with PVOD including no worsening with sildenafil and an absence of alveolar hemorrhage, the case was still thought to be due to PVOD despite the lack of a confirmatory lung biopsy. The diagnosis of PVOD was aided by the child's tendency toward pulmonary edema with hypoxemia, patchy infiltrates visualized on chest CT, and tricuspid regurgitation seen on ECHO. Jordan Holthe: Writing—original draft; conceptualization; investigation; writing—review and editing. Paul Boesch: Writing—review and editing; data curation; supervision. Mira Kohorst: Supervision; writing—review and editing. Jonathan Schwartz: Data curation; supervision; writing—review and editing. Asmaa Ferdjallah: Conceptualization; investigation; writing—original draft; writing—review and editing; supervision; data curation. The authors have no funding to report. The authors declare no conflict of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. At this point in time, a broad differential was considered. A collaborative differential was developed in conjunction with numerous pediatric specialties, including pulmonology, cardiology, and hematology and oncology. One of the things on the differential was pulmonary embolism due to the patient's hypoxemia and possible occurrence of a pulmonary hypertensive crisis. Another thing on the differential was pneumonia. The patient had previously been treated with cefepime for 10 days due to bacteremia and possible pneumonia. Therefore, it was a possibility that the pneumonia had reoccurred or never been treated completely in the first place. The leading diagnosis was pulmonary veno-occlusive disease. The team came to this conclusion based on the fact that infectious workup was negative and CT angiogram was negative, along with the patient's persistent hypoxemia and the tendency towards pulmonary edema with diuretic responsiveness. The team then decided to move forward with treatment. Lung biopsy was ultimately deferred due to the increased risk of anesthesia with the severe right ventricle dysfunction that the patient had. The cardiology and anesthesia teams were concerned about precipitating a hypertensive crisis and hemodynamic collapse leading to a prolonged intubation, and so the risks outweighed the potential benefit of a clear diagnosis and empiric treatment was started. The patient was started on defibrotide due to its known ability to treat VOD. Eventually the patient was additionally started on milrinone, epinephrine, and methylprednisolone. The team ordered repeat ECHOS to monitor for signs of improvement. The pathophysiology of pulmonary VOD is complex and remains incompletely understood. Therefore the treatment for this disease really required both cardiology and bone marrow support. From a cardiology standpoint, maximizing oxygen saturation was key, as was right ventricular optimization and improvement in systolic function with milrinone and epinephrine. The advantages of this approach are to improve organ function and address the core pathophysiology of essentially a weak heart. The disadvantages of this, of course, is that it is not addressing the exact pathophysiology of pulmonary VOD. From a bone marrow transplant standpoint, defibrotide was used as it has a very specific mechanism of action for VOD in general. Anticoagulation was deferred due to the high risk of bleeding and steroids were started to allow for anti-inflammatory properties as well as cytokine inhibition. Due to the high risk of using anesthesia in a patient with significant pulmonary hypertension, a diagnostic lung biopsy was not able to be performed and was deferred. Therefore, the clinical team had to rely on blood lab tests as well as imaging to guide understanding of what was occurring. From a cardiology standpoint, this includes a chest X-ray, ECHO, CT Angiogram, pro-BNP, and pulse ox to assess oxygenation. From a bone marrow transplant standpoint, blood pressure was studied intensely as was other broad blood tests referring to cell counts. For this patient, there were quite significant risk factors that contributed to endothelial injury and subsequently increased the likelihood of developing pulmonary VOD. Any patient who undergoes myeloablative chemotherapy is at risk for VOD. Particularly for this patient as part of his regimen, he underwent 3 back-to-back myeloablative stem cell transplants or rescues. So he was at an even increased risk of developing VOD in general as well as other endothelialopathies. This is always the case when patients undergo myeloablative chemotherapies as chemotherapy can cause significant damage to the microvasculature. This case is important because it shows that pulmonary VOD can actually be treated, and treated without a pulmonary biopsy. Lung biopsies can be very helpful in making a rapid and accurate clinical diagnosis; however, they are also very invasive and not always practical clinically. On paper, physicians want to make a clear diagnosis, however with a patient in front of us, the clinical status of the patient supersedes all else and we need to figure out the best way to pick an 'empiric treatment' without adding significant risk to the patient. Additionally, this case highlights the absolute non-negotiable importance of multidisciplinary care for all patients undergoing stem cell transplant. This patient had caring providers from bone marrow transplant team, cardiology, and pulmonology, who were willing to work together to come up with a unifying diagnosis.