A 32-year-old nulliparous pregnant woman at 21 weeks of gestation was referred to Kwong Wah Hospital in March 2021 for fetal right cystic lung mass (2.16×1.99×2.50 cm³). Repeat examination at 22 weeks of gestation revealed a right multicystic lung mass (3.46×2.46×2.37 cm³) with a dominant 2-cm cyst, suggestive of macrocystic congenital pulmonary airway malformation (CPAM). There was mild mediastinal shift but no hydrops. The CPAM volume ratio, calculated as (length×height×width×0.52)/head circumference, was 0.51. Amniocentesis with chromosomal microarray analysis showed no copy number variants. At 26 weeks of gestation, the lesion had enlarged to 5.73×4.28×4.26 cm³ (CVR=2.16), with mediastinal shift and mild ascites but no polyhydramnios. At 27 weeks of gestation, the lesion was dominated by a single cyst that measured 6.15×4.25×4.1 cm³ (CPAM volume ratio=2.09), with moderate polyhydramnios, ascites and skin oedema suggestive of fetal hydrops. Intramuscular betamethasone was administered for fetal lung maturation in view of the high risk of preterm delivery. Fetal thoracoamniotic shunting was offered to relieve the mass effect and fetal hydrops. The next day, a Somatex shunt (SOMATEX Medical Technologies, Berlin, Germany) was inserted into the CPAM under ultrasound guidance and 800 mL amniotic fluid was drained via the shunt cannula. Examination 9 days later showed CPAM with reduced size (2.3×1.7×1.5 cm³; CVR=0.12), shunt in situ, and no hydrops (Table).

At 29 weeks of gestation, the patient went into preterm prelabour with rupture of membrane that sealed off spontaneously. Serial ultrasound examinations showed satisfactory fetal growth, collapsed CPAM with shunt in situ, normal liquor volume and no hydrops (Table). Labour was induced at 38 weeks of gestation for oligohydramnios. A 2.75-kg female baby was delivered by vacuum extraction for maternal exhaustion, with paediatrician standby. The Apgar scores of the baby were 8 at 1 minute and 10 at 5 minutes and the arterial cord blood pH was 7.31, with base excess of -7.1 mmol/L. The shunt was specifically searched for immediately after delivery but the skin was intact (Fig a). The baby was given continuous positive airway pressure because of respiratory distress and was transferred to the neonatal intensive care unit. Urgent chest X-ray revealed the shunt in the right chest with right pneumothorax (Fig b). A chest drain was inserted and the baby was intubated. Computed tomography of the thorax of the baby on day 1 of life showed an irregular 4.2×3.5×2.5 cm³ cystic lesion in the right lung with the distal end of the shunt migrated between the chest wall and the scapular, abutting the right subscapularis muscle. Thoracoscopy on day 6 of life confirmed that one end of the shunt was within the CPAM in the right middle lobe, while the other end was at the subscapular space. The shunt was removed intact and near-total right middle lobe excision was performed thoracoscopically. The baby was successfully weaned off oxygen 3 weeks postoperatively and discharged 5 weeks later.

Congenital pulmonary airway malformations are uncommon lung lesions characterised by an overgrowth of terminal respiratory bronchioles that are often immature and non-functioning. A CPAM is considered macrocystic if at least one cyst is >5 mm and microcystic if the lesion appears echogenic on ultrasound examination. Although a microcystic CPAM may regress spontaneously after 26 to 28 weeks of gestation, most macrocystic CPAMs do not.1 Fetuses with large cystic lesions are also at risk of pulmonary hypoplasia and development of fetal hydrops due to compression of lung tissue and venous return. The survival of a hydropic fetus with congenital lung lesion has been reported to be 38%, compared with 87% for a fetus without hydrops.1 Studies have demonstrated a favourable outcome following thoracoamniotic shunting for macrocystic CPAM, with reduction in lesion volume, resolution of hydrops and improved survival.2 3 A systematic review showed an improved survival from 3% to 62% in hydropic fetuses treated with shunting.4 In a single-centre case series, survival was significantly associated with gestational age at birth, hydrops resolution and higher percent reduction in the size of the lung lesion following shunting.3

Although thoracoamniotic shunting improves fetal outcome, complications such as shunt occlusion, displacement, dislodgement, bleeding and chest wall deformation have been reported.3 5 6 7 Following successful drainage of the lesion, its surrounding normal lung parenchyma expands and grows. This may result in inward migration of the shunt. In a retrospective review,5 thoracoamniotic shunts inserted for primary pleural effusions and macrocystic CPAMs were antenatally displaced in 8.5% of fetuses, of which two-thirds migrated into the thorax. Re-shunting may be required if the displaced shunt fails to drain and fluid re-accumulates.5 Retained intrathoracic shunts may be managed conservatively as they are well tolerated without untoward postnatal sequalae.5 8 Nonetheless surgical removal may be necessary if the baby develops complications such as respiratory distress or tension pneumothorax.7 9

The Somatex shunt is commonly used to treat fetuses with obstructive urinary tract disorders. Recently, thoracoamniotic shunting with a Somatex shunt has been reported effective in relieving fetal pleural effusions with good survival rate although shunt dislodgement and entrapment has been reported in four of eight cases.10 Thoracoscopic removal of a displaced Somatex shunt has been reported necessary in a newborn with respiratory distress and progressive pleural effusion.7 In comparison with other commonly used shunts, such as Harrison and Rocket, Somatex insertion has multiple advantages including a finer introducer (1.2 mm) but a bigger shunt lumen (2.4 mm). It is also made of metal facilitating its easy identification antenatally on ultrasound or postnatally with X-rays or computed tomography.

A thoracoamniotic shunt should be clamped immediately following delivery to prevent air from entering the thorax and causing pneumothorax.2 In the current case, we did not expect air to enter the pleural cavity because the shunt was buried inside the skin of the baby. Our hypothesis is that air may have entered from the lung tissue into the pleural space via the displaced shunt. This is similar to the reported case of tension pneumothorax due to an internally displaced thoracoamniotic shunt communicating between the CPAM and the pleural cavity and diagnosed following neonatal resuscitation for apnoea.9 Both cases illustrate that pneumothorax is a possible and potentially life-threatening complication of an internally displaced shunt. It should be anticipated at birth and preparations made for emergency needle thoracocentesis. Obstetricians should be aware of the possible complications of thoracoamniotic shunts, and paediatricians should be alerted so that the newborn can receive prompt assessment and treatment.

Concept or design: VYT Chan, WC Leung, TY Leung.
Acquisition of data: VYT Chan, WT Tse, MC Chan, KKY Wong.
Analysis or interpretation of data: VYT Chan.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: KKY Wong, WC Leung, TY Leung.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

As an editor of the journal, KKY Wong was not involved in the peer review process. Other authors have disclosed no conflicts of interest.

This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

The patient and her baby were treated in accordance with the Declaration of Helsinki. Parental consent was obtained for the patient’s baby, and informed consent was obtained from the patient for all treatments and procedures, and publication of the case report.

1. Walker L, Cohen K, Rankin J, Crabbe D. Outcome of prenatally diagnosed congenital lung anomalies in the North of England: a review of 228 cases to aid in prenatal counselling. Prenat Diagn 2017;37:1001-7. Crossref

2. Schrey S, Kelly EN, Langer JC, et al. Fetal thoracoamniotic shunting for large macrocystic congenital cystic adenomatoid malformations of the lung. Ultrasound Obstet Gynecol 2012;39:515-20. Crossref

3. Peranteau WH, Adzick NS, Boelig MM, et al. Thoracoamniotic shunts for the management of fetal lung lesions and pleural effusions: a single-institution review and predictors of survival in 75 cases. J Pediatr Surg 2015;50:301-5. Crossref

4. Knox EM, Kilby MD, Martin WL, Khan KS. In-utero pulmonary drainage in the management of primary hydrothorax and congenital cystic lung lesion: a systematic review. Ultrasound Obstet Gynecol 2006;28:726-34. Crossref

5. Abbasi N, Windrim R, Keunen J, et al. Perinatal outcome in fetuses with dislodged thoraco-amniotic shunts. Fetal Diagn Ther 2021;48:430-9. Crossref

6. Makishi A, Kiyoshi K, Funakoshi T. EP21.22: Fetal chest wall deformity after thoracoamniotic shunting using a double-basket catheter for chylothorax: a case report. Ultrasound Obstet Gynecol 2016;48:362-3. Crossref

7. Sham GT, Chung PH, Chan IM, Leung WC, Wong KK. Thoracoscopic removal of a displaced thoracoamniotic shunt in a newborn with antenatal pleural effusion—a case report. Transl Pediatr 2020;9:702-6. Crossref

8. Tan AP, Tan B, Wright A, Kong JY. Management dilemma in thoracoamniotic shunt migrations. BMJ Case Rep 2023;16:e255760. Crossref

9. Law BH, Bratu I, Jain V, Landry MA. Refractory tension pneumothorax as a result of an internally displaced thoracoamniotic shunt in an infant with a congenital pulmonary airway malformation. BMJ Case Rep 2016:2016:bcr2016216324. Crossref

10. Chung MY, Leung WC, Tse WT, et al. The use of Somatex shunt for fetal pleural effusion: a cohort of 8 procedures. Fetal Diagn Ther 2021;48:440-7. Crossref

A 17-year-old male was referred to our institution in January 2016 due to elevated serum creatinine level of 1.85 mg/dL and nephrotic proteinuria level of 6839 mg/day. He had a history of epilepsy and had used various antiepileptic drugs (phenobarbital, valproic acid, and carbamazepine) from the ages 3 to 14 years. Physical examination revealed lower extremity oedema and a blood pressure of 140/90 mm Hg. Laboratory tests on admission showed a blood urea nitrogen level of 24 mg/dL, serum creatinine level of 1.68 mg/dL and a serum albumin level of 2.7 g/dL. Urine microscopy revealed three red blood cells per high-power field. A 24-hour urine collection revealed massive proteinuria level of 10.957 mg/day (296 mg/m²/h). Serum complement levels were normal and autoimmune tests (antinuclear antibodies, anti–double-stranded DNA antibodies, anti–glomerular basal membrane antibodies, and anti-neutrophil cytoplasmic antibodies) were negative. Viral serology, including hepatitis B virus, hepatitis C virus, human immunodeficiency virus, Epstein–Barr virus, cytomegalovirus, and parvovirus, was also negative. Abdominal ultrasound revealed increased echogenicity in the renal parenchyma.

Treatment with enalapril at a dose of 0.4 mg/kg/day was started. Kidney biopsy was performed on the seventh day after admission and showed compatibility with collapsing glomerulopathy (CG) [Fig 1]. Methylprednisolone boluses of 1 g were administered for 5 consecutive days, followed by oral prednisone at a dose of 60 mg/m²/day. By the 15th day, serum creatinine levels were at 2.1 mg/dL, serum albumin level at 1.9 g/dL, and 24-hour urine protein level at 17.8 g/day. Mycophenolate mofetil was added to the treatment regimen. On the 40th day, serum creatinine level had increased to 4.2 mg/dL with proteinuria of 13.3 g/day, leading to the initiation of rituximab and tapering of prednisolone. By the 60th day, mycophenolate mofetil was discontinued due to leukopenia; however, the patient had completed the 4 doses of weekly 375 mg/m²/dose rituximab treatment. Additionally, he received albumin infusions, diuretics, and antihypertensives. Since the clinical features and laboratory parameters did not improve, the patient underwent plasma exchange. After two sessions, haemodialysis was required due to worsening symptoms, uncontrolled hypervolaemia, and renal failure. No additional immunosuppressive was given at that time and the patient continued to receive haemodialysis. Genetic testing for mutations of the NPHS1 and NPHS2 genes were negative.

Another 17-year-old male was admitted to our institution in February 2016 for syncope. He had a history of headaches with intermittent vomiting for the previous 2 months and had been treated with metamizole, domperidone, zolmitriptan, and diclofenac. His mother had a history of minimal change disease aged 6 years. The patient’s blood pressure was measured as 200/120 mm Hg, and hypertensive retinopathy was observed during the ophthalmological examination. Initial serum creatinine level was 4.7 mg/dL and serum albumin level was 3.4 g/dL. Ferritin and parathyroid hormone levels were 274 ng/mL and 220 pg/mL, respectively. Microscopic urinalysis showed eight red blood cells per high-power field. He had nephrotic proteinuria of 3820 mg/day (91.5 mg/m²/h). Viral serology and autoimmune tests (antinuclear antibodies, anti–double-stranded DNA antibodies, anti–glomerular basal membrane antibodies, and anti-neutrophil cytoplasmic antibodies) were negative and complement levels were normal. Abdominal ultrasound revealed increased renal echogenicity. Cranial magnetic resonance imaging showed signs of posterior reversible encephalopathy syndrome. Hypertension was controlled using intravenous and oral antihypertensives (esmolol, captopril, amlodipine, doxazosin, and minoxidil). On the fourth day, the serum creatinine level increased to 5.9 mg/dL and the albumin level decreased to 2.4 g/dL. Kidney biopsy showed severe CG (Fig 2). Because the findings were chronic, no steroids or other immunosuppressive treatment were administered. Genetic testing for mutations of the NPHS1 and NPHS2 genes was negative. By the fifth month, the patient’s serum creatinine level had reached 6.9 mg/dL. After 1 year of peritoneal dialysis, he received a renal transplant.

Collapsing glomerulopathy is a histopathological pattern of podocytopathies.1 It was previously classified as a variant of focal segmental glomerulosclerosis (FSGS), known as collapsing FSGS.2 3 4 However, it is more severe at the initial stage and progresses more rapidly to end-stage kidney disease compared with non-collapsing FSGS, even when treatment is given.2 3 4 5 It typically presents with nephrotic proteinuria and elevated serum creatinine level, and is rare among children.3 5

Both patients had high serum creatinine level, nephrotic proteinuria, and hypertension. To establish the exact diagnosis and determine the prognosis, a kidney biopsy was performed as the gold standard for diagnosis. Histopathological findings of CG include glomerular capillary collapse in at least one glomerulus; hyperplasia and hypertrophy of visceral epithelial cells leading to pseudo-crescent formation; presence of periodic acid–Schiff-positive hyaline droplets in visceral epithelial cell cytoplasm; and severe tubulointerstitial inflammation in the early stages. Glomerulosclerosis and interstitial fibrosis are observed in the late stages, and immunofluorescence assay is typically negative.1 2 3 4 6 Kidney biopsies in both cases showed advanced CG with global glomerulosclerosis and interstitial fibrosis (Figs 1 and 2).

Collapsing glomerulopathy can be either idiopathic (primary), genetic (familial), or reactive (secondary).1 The idiopathic form is characterised by the loss of maturity markers and the re-expression of immaturity markers leading to the proliferation of immature podocytes.1 Secondary causes of CG include infections (human immunodeficiency virus, parvovirus B19, cytomegalovirus, hepatitis C virus, severe acute respiratory syndrome coronavirus 2), drugs (including valproic acid and anabolic steroids), autoimmune diseases (such as systemic lupus erythematosus), and malignancies.1 2 3 4 7 Genetic CG is associated with mitochondrial dysfunction that causes podocyte proliferation.1 Case 1 had a history of long-term use of antiepileptic drugs (phenobarbital, valproic acid, and carbamazepine). However, we found no other aetiological factors in either patient. Therefore, we concluded that while the aetiology in Case 1 could be idiopathic or valproic acid–related, it was idiopathic in Case 2.

There is no specific treatment for CG2; as such, the mainstay of therapy is for the disorders resulting from nephrotic syndrome (such as hypertension and oedema), treatment of the underlying conditions (such as infections and autoimmune diseases), and immunosuppressive therapy.8 Possible factors for progression to end-stage kidney disease in CG include a serum creatinine level >2 mg/dL at the time of biopsy, proteinuria >8 g/day and lack of remission, collapsing lesions in >20% of glomeruli, and the severe tubular changes and interstitial fibrosis.3 9 10 In Case 1, the rationale for aggressive immunosuppressive treatment was based on an initial serum creatinine level of 1.6 mg/dL, intense polymorphonuclear leukocytes and eosinophil infiltration, and 2 out of 24 glomeruli showing glomerulosclerosis. Case 2 did not receive immunosuppressive treatment due to the chronicity of the disease and advanced global glomerulosclerosis (67%). The Table summarises the clinical findings in both patients.

It is important to recognise that CG is a separate clinicopathological entity from FSGS. Due to the poor response to immunosuppressive drugs and the potential for renal transplantation, we recommend avoiding aggressive immunosuppressive therapy for patients with poor prognostic factors at the time of diagnosis. This approach helps minimise the side-effects of cumulative immunosuppression.

Concept or design: All authors.
Acquisition of data: All authors.
Analysis or interpretation of data: All authors.
Drafting of the manuscript: Y Özdemir Atikel, SA Bakkaloğlu.
Critical revision of the manuscript for important intellectual content: All authors.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

All authors have disclosed no conflicts of interest.

The two cases have been presented during oral presentations at the IPNA Teaching Course of the 5th Southeastern Europe Pediatric Nephrology Working Group Meeting inSkopje, Macedonia, 10-11 June 2016.

This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Both patients were treated in accordance with the Declaration of Helsinki. Written informed consent for publication was obtained from both patients and their parents.

1. Barisoni L, Schnaper HW, Kopp JB. A proposed taxonomy for the podocytopathies: a reassessment of the primary nephrotic diseases. Clin J Am Soc Nephrol 2007;2:529-42. Crossref

2. Albaqumi M, Soos TJ, Barisoni L, Nelson PJ. Collapsing glomerulopathy. J Am Soc Nephrol 2006;17:2854-63. Crossref

3. Mubarak M. Collapsing focal segmental glomerulosclerosis: current concepts. World J Nephrol 2012;1:35-42. Crossref

4. Ferreira AC, Carvalho D, Carvalho F, Galvão MJ, Nolasco F. Collapsing glomerulopathy in Portugal: a review of the histological and clinical findings in HIV and non-HIV patients. Nephrol Dial Transplant 2011;26:2209-15. Crossref

5. Gulati A, Sharma A, Hari P, Dinda AK, Bagga A. Idiopathic collapsing glomerulopathy in children. Clin Exp Nephrol 2008;12:348-53. Crossref

6. Fogo AB, Lusco MA, Najafian B, Alpers CE. AJKD Atlas of Renal Pathology: collapsing glomerulopathy. Am J Kidney Dis 2015;66:e3-4. Crossref

7. Nasr SH, Kopp JB. COVID-19–associated collapsing glomerulopathy: an emerging entity. Kidney Int Rep 2020;5:759-61. Crossref

8. Cutrim ÉM, Neves PD, Campos MA, et al. Collapsing glomerulopathy: a review by the Collapsing Brazilian Consortium. Front Med (Lausanne) 2022;9:846173. Crossref

9. Laurinavicius A, Hurwitz S, Rennke HG. Collapsing glomerulopathy in HIV and non-HIV patients: a clinicopathological and follow-up study. Kidney Int 1999;56:2203-13. Crossref

10. Valeri A, Barisoni L, Appel GB, Seigle R, D’Agati V. Idiopathic collapsing focal segmental glomerulosclerosis: a clinicopathologic study. Kidney Int 1996;50:1734-46. Crossref

A 70-year-old Chinese woman was diagnosed with stage IV anaplastic lymphoma kinase (ALK)–positive non–small-cell lung carcinoma with intrapulmonary, pleural and lymph node metastasis in December 2022. Her private oncologist first prescribed lorlatinib in mid-December 2022. The drug had been well tolerated by the patient and there was no evidence of myelotoxicity.

The patient attended a public hospital for further care in February 2023. Lorlatinib was switched to alectinib on 17 March 2023. Her baseline haemoglobin level was 11.1 g/dL on 16 March 2023 prior to commencement of alectinib. Subsequent follow-ups on 13 April and 3 May revealed that her haemoglobin level had fallen to 10 g/dL and 8.7 g/dL, respectively (Fig 1a). Mean cell volume was 65.4 fL. She had no clinical signs or symptoms suggestive of acute blood loss.

Peripheral blood smear showed marked red cell sphero-acanthocytosis (Fig 2). The previously normal bilirubin level rose to 43 μmol/L and lactate dehydrogenase (LDH) level to 317 IU/L. Haptoglobin level was very low (<0.06 g/L) [Fig 1b]. Direct Coombs test (DAT) was negative. Alectinib had been withheld since 3 May 2023 in view of the potential differential diagnosis of drug-induced haemolytic anaemia. After suspending alectinib, the haemoglobin level remained static for 1 week (8.7 g/dL on 10 May 2023). It then increased slowly over the following week (9.1 g/dL on 15 May 2023) [Fig 1a]. Lactate dehydrogenase level also decreased to 278 IU/L, with total bilirubin level decreased to 24 μmol/L and direct bilirubin level normalised (Fig 1b).

Since the anaemia of the patient had improved, another ALK inhibitor, brigatinib, was prescribed on 18 May 2023, at 90 mg daily. Her haemoglobin level increased to 9.7 g/dL after 2 weeks. Total bilirubin level was normalised on 1 June 2023 (Fig 1a). Brigatinib was then stepped up to 180 mg daily (full dose) as it was well tolerated.

The patient demonstrated a favourable response to ALK inhibitors as evidenced by a continuously decreasing cancer embryonic antigen level. A contrast computed tomography scan of the thorax, abdomen and pelvis on 30 May 2023 showed partial response after 5 months of lorlatinib and alectinib. The primary lung tumour at the right lower lobe showed a partial response, decreasing from 6.2 cm to 1.5 cm in size. There was no evidence of bone metastasis.

Anaemia caused by alectinib is uncommon although clinical trials have reported clinically significant anaemia (grade ≥3) in around 7% of cases.1 Nonetheless there are limited reports of alectinib-induced haemolytic anaemia.2 Misawa et al3 described a case of grade 4 anaemia due to drug-induced haemolytic anaemia in 2023. No such case has been reported in Hong Kong to date.

The index patient demonstrated alectinib-induced haemolytic anaemia with morphological change to erythrocytes and negative DAT result. The blood smear of sphero-acanthocytosis combined with altered bilirubin, LDH and haptoglobin was strongly suggestive of haemolytic anaemia. The improvement in haemoglobin, bilirubin and LDH levels following suspension of alectinib suggested that the haemolytic anaemia was drug induced.

Drug-induced haemolytic anaemia is usually due to drug-induced immune haemolytic anaemia.4 Nonetheless in this case, an autoimmune cause was excluded due to the patient’s negative DAT result even though 5% to 10% of DAT-negative cases may have an immune component.2 Differential diagnoses of DAT-negative haemolytic anaemia include membranopathies (eg, hereditary spherocytosis), thrombotic microangiopathies (eg, thrombotic thrombocytopenic purpura), enzymopathies (eg, glucose-6-phosphate dehydrogenase), infection (eg, malaria or Clostridium), and haemoglobinopathies (eg, sickle cell disease).5 Our patient had no signs of infection or history of haematological disease. The temporal relationship between administration of alectinib and the occurrence of haemolytic anaemia favoured a diagnosis of drug-induced haemolysis. The laboratory findings were also consistent with other case reports.6

The precise mechanism is uncertain. It has been postulated that alectinib induces erythrocyte membrane changes.3 The presence of spherocytes in the peripheral blood film may arise from these membrane changes (Fig 2). Additional investigation is warranted to further understand the underlying mechanism.

Our patient developed haemolytic anaemia within 2 months of commencing alectinib. Misawa et al3 reported that grade 4 haemolytic anaemia could occur after 3 years. Regular monitoring of haemoglobin should be undertaken in patients prescribed alectinib. Haemolytic anaemia workup, including peripheral blood smear, bilirubin, haptoglobin and LDH levels, should be considered if indicated.

This case demonstrated no cross reactivity among other ALK-positive first-line targeted therapies (lorlatinib and brigatinib). Our patient first commenced lorlatinib following a private consultation and later switched to alectinib with funding support. The patient continues her treatment with brigatinib. There was no documented drop in haemoglobin level after lorlatinib, and following discontinuation of alectinib and initiation of brigatinib, her haemoglobin level showed an improving trend. Limited case reports of alectinib-induced haemolytic anaemia have been managed by discontinuation of therapy or rechallenge with alectinib at a reduced dosage.2 3 6 To the best of our knowledge, cross reactivity among first-line ALK tyrosine inhibitors (lorlatinib, alectinib and brigatinib) has not been reported. Our case demonstrates that it is safe to switch treatment to an alternative ALK inhibitor when alectinib-induced haemolytic anaemia occurs. The drug-induced haemolytic anaemia is specific to alectinib.

This case highlights the importance of interval haemoglobin monitoring. A persistent drop in haemoglobin level following initiation of alectinib warrants prompt investigations for possible differential diagnosis of alectinib-induced haemolytic anaemia. This case also suggests that it is safe to switch to an alternative ALK inhibitor in the presence of alectinib-induced haemolytic anaemia.

Both authors contributed to the concept or design of the study, acquisition of the data, analysis or interpretation of the data, drafting of the manuscript, and critical revision of the manuscript for important intellectual content. Both authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

Both authors have disclosed no conflicts of interest.

The authors thank Dr Kenneth Mung from the Department of Pathology of Pamela Youde Nethersole Eastern Hospital for providing microscopic photos of the blood smear of the patient.

This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

The patient was treated in accordance with the Declaration of Helsinki. The patient provided consent for all treatments and procures, and consent for publication of this case report.

1. Mok T, Camidge DR, Gadgeel SM, et al. Updated overall survival and final progression-free survival data for patients with treatment-naive advanced ALK-positive non–small-cell lung cancer in the ALEX study. Ann Oncol 2020;31:1056-64. Crossref

2. Okumoto J, Sakamoto S, Masuda T, et al. Alectinib-induced immune hemolytic anemia in a patient with lung adenocarcinoma. Intern Med 2021;60:611-5. Crossref

3. Misawa, K, Nakamichi S, Iida H, et al. Alectinib-induced severe hemolytic anemia in a patient with ALK-positive non–small cell lung cancer: a case report. Onco Targets Ther 2023;16:65-9. Crossref

4. Garbe E, Andersohn F, Bronder E, et al. Drug induced immune haemolytic anaemia in the Berlin Case-Control Surveillance Study. Br J Haematol 2011;154:644-53. Crossref

5. Palmer D, Seviar D. How to approach haemolysis: haemolytic anaemia for the general physician. Clin Med (Lond) 2022;22:210-3. Crossref

6. Gullapalli V, Xu W, Lewis CR, Anazodo A, Gerber GK. A multi-centre case series of alectinib-related erythrocyte membrane changes and associated haemolysis. J Hematop 2021;14:131-6. Crossref

A 58-year-old Chinese woman with good past health presented with a 2-day history of right lower abdominal pain. Routine biochemical blood tests revealed a white blood cell count of 9.1 × 10⁹/L and normal renal function (creatinine level: 49 μmol/L). An urgent contrast-enhanced computed tomography scan revealed acute diverticulitis in the caecum. Incidentally, a solitary well-defined homogenous fat-density mass with unenhanced density of -63 Hounsfield units was evident at the right renal sinus, with extension into the right renal vein and inferior vena cava (IVC) [Fig 1]. Findings were suggestive of an aggressive renal angiomyolipoma (AML) with right renal vein and IVC thrombus. The patient was referred to an urologist. Subsequent robotic-assisted laparoscopic radical nephrectomy and IVC thrombectomy were performed uneventfully.

The surgical specimen provided by the pathologist demonstrated the right renal vein depicted as a purple red strip and fatty tumour (Fig 2a). Microscopic examination revealed all features of a typical AML including mature adipocytes, myoid spindle cells, and thick-walled blood vessels (Fig 2b). It was also positive on Melan A (Fig 2c) and HMB-45 stains (Fig 2d).

Angiomyolipoma is a common benign renal mesenchymal neoplasm composed in variable proportions of adipose tissue, smooth muscle cells, and abnormal vessels. It is most commonly sporadic (80%) or associated with genetic syndromes such as tuberous sclerosis.1 A renal mass with visualisation of macroscopic fat density on computed tomography is virtually diagnostic of AML. Therefore, visualisation of fat density in the intravascular compartment is indicative of tumour invasion.

Cases of aggressive AML with renal vein and IVC thrombus have been reported. A literature review2 of 26 cases of invasive renal AML reported that a large size and right-sided location of the tumour may be contributing factors in AML with intravascular invasion. This may be related to the shorter and straighter course of the right renal vein. Two variants of renal AML, namely, classic and epithelioid, have been described. The epithelioid variant has been reported to exhibit aggressive behaviour with IVC thrombus.3 In this case, the patient had a classic subtype of renal AML with mature adipocytes (Fig 2b) and no epithelioid cells.

The optimal treatment is robotic nephrectomy and IVC thrombectomy irrespective of tumour size, since IVC thrombus can be life threatening with higher risks of vessel occlusion and tumour embolus. A multi-institutional study in 2016 retrospectively reviewed 32 cases of robotic radical nephrectomy and IVC thrombectomy and demonstrated a favourable outcome with adequate robotic experience including less blood loss, earlier patient discharge, and fewer complications compared with open nephrectomy.4 Inferior vena cava thrombectomy follows similar surgical principles to that for renal cell carcinoma with IVC thrombus. Proper visualisation and mobilisation of the kidney, renal vein and IVC allow selective vascular clamping, namely, the infrarenal and suprarenal IVC as well as the renal vein. The IVC is then exposed and the tumour is dissected away from the IVC. The cavotomy is then closed and nephrectomy is performed.

We describe a classic subtype of renal AML extending into the renal vein and IVC, managed by robotic-assisted laparoscopic nephrectomy and IVC thrombectomy. Computed tomography and other imaging modalities are essential to diagnose AML and for proper surgical planning. With proper visualisation of the kidney, renal vein and IVC, a safe and successful outcome of robotic nephrectomy and IVC thrombectomy is achieved.

Concept or design: PHW Lo, HM Kwok, FH Ng, JKF Ma.
Acquisition of data: PHW Lo, HM Kwok, FH Ng, HY Lo.
Analysis or interpretation of data: PHW Lo, HM Kwok, HY Lo.
Drafting of the manuscript: All authors.
Critical revision of the manuscript for important intellectual content: All authors.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

All authors have disclosed no conflicts of interest.

This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

This study was approved by the Kowloon West Cluster Research Ethics Committee of Hospital Authority, Hong Kong [Ref No.: KW/EX-23-010(180-06)]. The patient has given a consent statement for publication of this case report.

1. Martignoni G, Amin M.B. Pathology and genetics of tumours of the urinary system and male genital organs. In: Ebie JN, Sauter G, Epstein JI, editors. World Health Organization Classification of Tumours. Lyon, France: IARC; 2004: 65-7.

2. Islam AH, Ehara T, Kato H, Hayama M, Kashiwabara T, Nishizawa O. Angiomyolipoma of kidney involving the inferior vena cava. Int J Urol 2004;11:897-902. Crossref

3. Fox C, Salami SS, Moreira DM, et al. Aggressive renal angiomyolipoma of the lipomatous variant with inferior vena cava thrombus: a case report and review of the literature. Urol Case Rep 2013;2:9-11. Crossref

4. Abaza R, Shabsigh A, Castle E, et al. Multi-institutional experience with robotic nephrectomy with inferior vena cava tumor thrombectomy. J Urol 2016;195:865-71. Crossref

We report the first case of atypical fifth metatarsal fracture in a 69-year-old Chinese woman following long-term treatment for osteoporosis with denosumab. She had been diagnosed with hypertension in 2007 and was managed with amlodipine and losartan. One year previously in 2006, she had been diagnosed with postmenopausal osteoporosis based on low bone mineral density (BMD) revealed at a health screening. Her height and weight were 150 cm and 54 kg, respectively, with a body mass index of 24.0 kg/m². She had no previous history of fracture or parental history of hip fracture.

The patient was initially prescribed raloxifene for 2 years but due to a suboptimal BMD response, she was treated with alendronate for 4 years. Given the prolonged use of alendronate, she was switched to strontium ranelate but this was discontinued a few months later due to intolerance. Dual-energy X-ray absorptiometry at that juncture in 2014 showed the BMD of the L1-L4 level, the left femoral neck, and the left total hip were 0.928 g/cm² (T-score=-2.1), 0.646 g/cm² (T-score=-2.9), and 0.725 g/cm² (T-score=-2.4), respectively. She was started on subcutaneous denosumab 60 mg every 6 months that was continued until October 2021. Her most recent dual-energy X-ray absorptiometry in December 2020 revealed the BMD of the left femoral neck and left total hip were 0.544 g/cm² (T-score=-2.3) and 0.715 g/cm² (T-score=-1.4), respectively, indicating significant improvement. The BMD of the lumbar spine was not reported due to significant degenerative changes.

The patient presented to the primary care physician in November 2021 with a 2-week history of mechanical right lateral foot pain. She reported no preceding injury but physical examination revealed mild tenderness over the base of the right fifth metatarsal. There was no joint effusion or signs of infection and distal neurovascular status was intact. X-ray of the right foot was performed (Fig a) and she was prescribed analgesics for right foot tendinitis. One month later in December 2021, she complained of persistent right foot pain during her scheduled follow-up at the Osteoporosis Centre of The University of Hong Kong. Physical examination revealed intense tenderness over the lateral side of her right foot, near the base of the fifth metatarsal. Review of the X-ray taken in November 2021 identified beaking over the lateral aspect at the base of the right fifth metatarsal, indicating a periosteal reaction, associated with a transverse radiolucency across the cortex. Repeat X-ray of the right foot (Fig b) revealed a more prominent transverse radiolucency across the same site. A diagnosis of denosumab-related atypical metatarsal fracture was reached given the long duration of antiresorptive treatment (likely related to denosumab since fracture occurred after 7 years of denosumab), prodromal symptoms and absence of high-energy trauma, along with characteristic radiological findings. Blood tests showed serum calcium level of 2.44 mmol/L (reference range, 2.24-2.63), phosphate level of 0.98 mmol/L (reference range, 0.88-1.45), creatinine level of 60 μmol/L (reference range, 49-82), parathyroid hormone level of 1.7 pmol/L (reference range, 1.3-6.8), and 25-hydroxyvitamin level of 70 nmol/L (sufficient level: 50-220). Although bone turnover markers were not measured, serum alkaline phosphatase level was on the low side at 48 U/L (reference range, 47-124) that could suggest suppressed bone turnover. An X-ray of the left foot was not available, but the patient had no history of left foot pain. For her atypical right metatarsal fracture, she was given a short leg plaster and completed a course of physiotherapy. The fracture healed completely in 6 months. She was started on teriparatide as anti-osteoporosis treatment, and this was tolerated well.

The American Society for Bone and Mineral Research Task Force published a report in 2014 on the case definition of atypical femoral fracture (AFF),1 which refers to fracture located along the femoral diaphysis just distal to lesser trochanter to just proximal to the supracondylar flare. Major and minor diagnostic criteria were proposed. The pathophysiology is postulated to be related to oversuppression of bone turnover impairing normal bone remodelling in response to physiological stress. Prolonged bisphosphonate use has been associated with rare adverse events of atypical fractures (3.2-50 cases per 100 000 person-years).1 Atypical femoral fractures related to denosumab use have been reported, but they are even rarer (0.8 per 10000 person-years).2 Our patient had been taking alendronate for 4 years and denosumab for 7 years. It was very likely that bone turnover had been severely suppressed. In line with this, her serum alkaline phosphatase level was at the lower end of the reference range.

Less commonly, atypical fractures involving long bones other than the femur have also been reported among patients prescribed bisphosphonate or denosumab, including the tibia and ulnar.3 Notably, atypical metatarsal fractures associated with bisphosphonate use for osteoporosis have been reported in only three cases.3 4 5 Our case is the first of atypical metatarsal fracture associated with denosumab. Atypical metatarsal fractures associated with antiresorptive agents, as illustrated in our patient and other case reports (Table), share several characteristics and are distinct from typical fifth metatarsal fractures. They were all associated with prodromal pain with no history of high-energy impact. More importantly, the radiological features fulfil some of the criteria of AFF proposed in the American Society for Bone and Mineral Research 2013 consensus,1 including a fracture line originating in the lateral cortex and periosteal reaction. Although the atypical metatarsal fracture in our patient was temporally related to the long duration of denosumab, the possibility that alendronate had caused abnormal microarchitecture that persisted and was perpetuated by denosumab therapy could not be excluded since she had also been prescribed alendronate for 4 years.

When AFF is diagnosed, antiresorptive agents should be stopped. The risk of AFF of the contralateral femur declines after discontinuation of bisphosphonate. Adequate vitamin D, calcium supplementation and teriparatide can promote healing and reunion of atypical femoral fracture. Surgery promotes healing and relieves pain for AFF. Incomplete AFF has been successfully treated with prophylactic intramedullary nailing. Unlike AFF, the treatment for atypical metatarsal fractures is less well defined. The decision about treatment involves a patient-physician discussion, although the presence of a longer transverse radiolucent line across the shaft favours surgical management.

Although bisphosphonate and denosumab are safe and efficacious, rare side-effects of atypical fractures have been reported. Our case highlights the importance of clinician vigilance for atypical fractures when patients on long-term antiresorptive agents complain of pain over the long bones (not limited to the femur) to enable timely management to reduce morbidities and re-evaluate the anti-osteoporosis strategy.

Concept or design: All authors.
Acquisition of data: EKH Leung, AKC Kan, YC Woo, DTW Lui.
Analysis or interpretation of data: EKH Leung, AKC Kan, SCW Cheung, YC Woo, DTW Lui.
Drafting of the manuscript: EKH Leung, AKC Kan, YC Woo, DTW Lui.
Critical revision of the manuscript for important intellectual content: All authors.

All authors had full access to the data, contributed to the study, approved the final version for publication, and take responsibility for its accuracy and integrity.

All authors have disclosed no conflicts of interest.

This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

The patient was treated in accordance with the Declaration of Helsinki. Informed consent was obtained from the patient for publication of this case report and any accompanying images.

1. Shane E, Burr D, Abrahamsen B, et al. Atypical subtrochanteric and diaphyseal femoral fractures: second report of a task force of the American Society for Bone and Mineral Research. J Bone Miner Res 2014;29:1-23. Crossref

2. Paparodis R, Buehring B, Pelley EM, Binkley N. A case of an unusual subtrochanteric fracture in a patient receiving denosumab. Endocr Pract 2013;19:e64-8. Crossref

3. Min BC, Chung CY, Park MS, Sung KH, Lee KM. A suspicious atypical fracture of 5th metatarsal bone: a case report. J Orthop Sci 2022;27:281-3. Crossref

4. Pradhan P, Saxena V, Yadav A, Mehrotra V. Atypical metatarsal fracture in a patient on long term bisphosphonate therapy. Indian J Orthop 2012;46:589-92. Crossref

5. Valiollahi B, Salehpour M, Bashari H, Majdi Sh, Mohammadpour M. Atypical metatarsal fracture in a patient on long-term bisphosphonate therapy (case report). J Res Orthopedic Sci 2019;6:25-30. Crossref