Type B lactic acidosis due to Warburg effect in a child presenting with T cell acute lymphoblastic leukaemia: a milder phenotype

Abstract

Lactic acidosis (LA) is characterised by persistently increased blood lactate >5 mmol/L. Type A LA due to anaerobic glycolysis is frequently seen during management of haematological malignancies. A rare form of LA known as type B LA, which occurs as a result of metabolic dysregulation at cellular level has been described recently. This has been reported to be because of Warburg effect (WE) or aerobic glycolysis, which is seen in cancerous cells as they rely on aerobic glycolysis rather than oxidative phosphorylation for energy generation. Presence of type B LA at initial presentation of haematological malignancies is a poor prognosticating factor and has rarely been reported in children. We present a child with T cell acute lymphoblastic leukaemia with mild phenotype of type B LA due to WE. She responded dramatically to definitive chemotherapy and tolerated intensive phase of chemotherapy without any significant morbidity.

Background

Lactic acidosis (LA) is characterised by persistently increased blood lactate >5 mmol/L in association with metabolic acidosis.¹ Type A LA is a common form of LA secondary to hypovolaemia, hypoxia, circulatory failure, pulmonary oedema, sepsis or regional tissue ischaemia and commonly complicate management of haematological malignancies. In anaerobic conditions as described above, the energy production is from anaerobic glycolysis leading to lactate production. Lately, a rare form of type B LA has been described, which is a result of metabolic dysregulation at cellular level rather than organ dysfunction.¹ The Warburg effect (WE) is a phenomenon seen in type B LA, which is exclusively seen in cancer cells as they rely on aerobic glycolysis rather than oxidative phosphorylation for energy production.² Type B LA at initial presentation of haematological malignancies has rarely been reported in children and is described to be a poor prognostic factor.^{1 3} We present a child with T cell acute lymphoblastic leukaemia (T -ALL) who had a mild phenotype of type B LA due to WE at initial presentation. She responded to definitive chemotherapy and also tolerated intensive phase of chemotherapy without any significant morbidity.

Case presentation

An 11-year-old girl born to a non-consanguineous marriage presented with history of unquantified weight loss with loss of appetite for last 3 months. During this period, she also had history of nausea and vomiting. The parents also gave history of pallor with worsening respiratory distress for last 10 days. There was no associated history of fever, cough, bleeding manifestation, blood loss, jaundice, bone pain, any blood component transfusion, rashes, diarrhoea or any urinary complaints.

On examination, she had severe pallor with no icterus or cyanosis. She had tachycardia and tachypnoea (heart rate—140/min and respiratory rate 35/min) with normal blood pressure and SPO₂. She had severe wasting with weight 19 kg (−4Z) and height 135 cm (−1.6Z). There were no features of congestive cardiac failure. She did not have lymphadenopathy or hepatosplenomegaly. Other system examination was normal.

Investigations

The investigations at initial presentation of this child are as summarised in table 1.

Table 1.

Investigations at initial presentation of index child

Test	Results	Normal range
Complete cell count
Hb (g/L)	36	117–157
Total leucocyte count (×109/L)	4.6	4–11
Differential leucocyte count	Neutrophils: 42%, lymphocytes: 55%, monocytes: 1%, eosinophils: 2%
Platelet count (×109/L)	160	150–450
Mean corpuscular volume (fL)	97	78–100
Mean corpuscular Hb (pg)	34	26.5–33
Reticulocyte count corrected (%)	1	<2
Peripheral blood film	Dimorphic picture with adequate platelets, no immature cells/blasts and no features of hemolysis
Biochemistry
Blood urea (mmol/L)	5.7	1.7–8.5
Serum creatinine (µmol/L)	30.4	25–79
Sodium (mmol/L)	134	133–144
Potassium(mmol/L)	2.4	3.4–5.3
Bilirubin (µmol/L)	13.6	3.4–22.2
Aspartate transaminase (U/L)	63	0–45
Alanine transferase (U/L)	36	0–50
Lactate dehydrogenase (U/L)	986	240–480
Random blood sugar (mmol/L)	5.8	2.2–7.7
Investigations for cause of anaemia
Direct Coombs test	Negative
Serum folate (nmol/L)	5.42	>12.2
Vitamin B12 (pmol/L)	575	182–672
Serum ferritin (pmol/L)	423	22–269
% transferring saturation	96%	>16%
Hb electrophoresis	HbA 97% and HbA2—2.9%
Investigations for cause of severe wasting with lactic acidosis
Venous blood gas
pH	7.11	7.35–7.45
pO2(mm Hg)	51.2
pCO2(mmHg)	23.1	36–44
HCO3 (mmol/L)	7.9	21–28
Lactate (mmol/L)	20	1–1.7
Urine examination	Protein—nil, RBC—nil, casts—nil
Spot protein/creatinine	0.83	<1
Spot urine potassium (meq/L)	16.6	<20
Spot urinary calcium/creatinine	0.1	<0.14
Tissue transglutaminase IgA (U/mL)	0.227	<7
Immunoglobulin A (g/L)	1.4	0.7–4
Miscellaneous
Ultrasonography abdomen	Normal size kidneys, normal echotexture of liver and kidneys
Blood culture	Sterile
ECG	No changes of hypokalemia
X-ray chest	Normal

ECG, Electrocardiograph; Hb, haemoglobin.

Differential diagnosis

In view of severe metabolic acidosis with LA and failure to thrive with normocytic anaemia she was worked up for chronic kidney disease initially. However repeat cell counts also showed thrombocytopaenia (platelet count—60×10⁹/L). In view of bicytopaenia a bone marrow infiltrative pathology was strongly considered and a bone marrow aspiration (BMA) was done which showed 90% blasts. Flowcytometry of BMA confirmed T-ALL; which was CD3+, cytoCD3+, CD4+, CD5+, CD7+, CD8+, CD10+(aberrant), CD34+ and TdT negative. Her karyotype showed normal female component.

Treatment

At admission, in view of severe anaemia she was transfused with packed red cells (30 mL/kg) in 3 aliquots over 3 days. Her tachycardia settled after transfusions; however, respiratory rate continued to be around 30/min. After initial investigations, in view of severe metabolic acidosis she was started on sodium bicarbonate correction along with intravenous maintenance fluid.

After confirmation of diagnosis of T-ALL she was started on hydration with 0.5 normal saline at 125 mL/m²/hour (without sodium bicarbonate) and oral allopurinol. After taking due consent from the parents, she was started on Berlin-Frankfurt-Munster (BFM) chemotherapy protocol (version 2002) for ALL. She was given oral prednisolone at 60 mg/m²/day for 4 weeks. Hydration was stopped on day 5 of steroids. She subsequently received four cycles of injectable daunorubicin at 30 mg/m² and vincristine 1.5 mg/m² weekly. She was also given eight doses of injectable L-asparginase and three doses of intrathecal methotrexate 12.5 mg as part of induction therapy.

Outcome and follow-up

The serial venous blood gases of index child from the day of admission till resolution of LA are depicted in table 2.

Table 2.

Serial venous blood gases of index child from the day of admission until resolution of lactic acidosis

	Normal range	D-3	D0	D+1	D+3	D+4	D+8	D+14
pH	7.35–7.45	7.11	7.15	7.18	7.42	7.44	7.439	7.36
pCO2 (mm Hg)	36–44	23.1	23.6	25.6	29.9	32.5	28.1	44.5
pO2 (mm Hg)		51.2	49.9	28.6	57.9	58.8	54.4	41
Lactate (mmol/L)	1–1.7	20	14.6	11	6.6	5.6	3.9	1.29
HCO3 (mmol/L)	21–28	7.9	8.4	9.7	19.6	22.5	19.2	25.2
Chloride (mmol/L)	96–106	105	107	111	106	101	112	114
Potassium (mmol/L)	3.4–5.3	2.55	2.54	2.38	2.12	2.59	3.91	3.9
Base excess(mmol/L)	−2 to +2	−19.1	−17.9	−16.4	−3.5	−0.6	−3.8
Anion gap	12±4	26	25	20.4	15.9	12.4	18.3

D—days from start of definitive chemotherapy.

LA was gradually corrected with start of steroids as part of induction therapy for T-ALL and completely normalised at end of 2 weeks of therapy. Her appetite improved and her weight increased to 24 kg (−2.5Z) at end of week 5 induction therapy.

She had a favourable response to induction chemotherapy as evident by a good steroid response at end of day 7 of therapy and at end of week 5 minimal residual disease by flowcytometry which was <0.01%. She was continued on BFM protocol and has completed the intensive phase of chemotherapy. Presently the child is on maintenance phase and is tolerating the chemotherapy well.

Discussion

LA is rarely associated with haematological malignancies like ALL. The aetiology of LA in ALL could be multifactorial and is mainly seen due to tissue hypoperfusion. However, type B LA due to WE is described primarily in association with haematological malignancies.^{1 2} The mechanism behind WE remains largely unexplained. The postulated explanation behind WE is first, a spillover effect of cell proliferation signalling pathways on metabolic pathways and second, cancer-associated mutations switching the cellular mechanism to proliferation rather than efficient energy production.² Some recent studies suggest a cause and effect relationship of WE with malignancies. It is hypothesised that cellular switch to aerobic glycolysis may represent the very time point of cancerous transformation of a normal cell.⁴

Type B LA due to WE is a diagnosis of exclusion. The diagnosis becomes more difficult in sick children as they are more likely to have type A LA due to tissue hypoperfusion. Clinical examination is non-specific and may reflect findings associated with metabolic acidosis. It is suspected in patients presenting with LA and hypoglycaemia, though the latter is not a persistent metabolic association. In a study of six cases on haematological malignancies with type B LA, only one-third had hypoglycaemia.³ Hepatic infiltration by malignant cells may explain association with hypoglycaemia.¹ The involvement of insulin-like growth factors in a bid to correlate LA with hypoglycaemia remains speculative.¹ The correction of metabolic abnormalities on starting therapy for the haematological malignancy is indirect evidence favouring type B LA. In our patient, organ dysfunction-related LA was ruled out. The child’s response to induction chemotherapy with complete resolution of LA favours type B LA. Hepatic infiltration by malignant cells was also not considered likely in view of absence of hypoglycaemia and normal liver size and echo-texture on ultrasonography in our patient. Liver and renal biopsy were not done as there was no evidence of their involvement on non-invasive tests and posed significant risk.

Early initiation of chemotherapy is the backbone of management of malignancy associated type B LA.^{1 3} The use of exchange buffer like sodium bicarbonate to decrease arterial pH in an attempt to diminish deleterious effect of acidosis have been questioned. Intracellular acidosis may slow lactate production and alkalanisation with sodium bicarbonate may potentiate lactate production and lead to fluid overload; therefore, may be counterproductive.⁵ Renal replacement therapy in form of dialysis may not sufficiently remove accumulated lactate.^{6 7} Thiamine is a pyruvate dehydrogenase enzyme cofactor which promotes aerobic glycosylation pathway and is recommended by some in the management of LA.⁶ Therefore, the initiation of definitive chemotherapy should take precedence over other non-established therapeutic interventions.

WE is commonly described in malignancies with high tumour burden and increased proliferation rates.⁸ In largest reported series of LA and haematological malignancies more than 80% of cases had high TLC.¹ In tumour cells, overexpression of glycolytic enzymes leads to accumulation of lactic acid which is directly proportional to cell burden.⁹ Also, high turnover rate of malignant haematological cells leads to hypoxic changes inside bone marrow triggering the WE.¹ The index child presented with low tumour burden as evident by low TLC and absence of significant hepatosplenomegaly and lymphadenopathy.

WE associated with haematological malignancies is an extremely poor prognostic marker.¹⁰ Most of the case series and reports reported a dismal outcomes in form of death or early relapse.1 3 10 11 In the largest reported series of 25 patients of LA due to all causes in leukaemia, 21/25 died within 2 weeks of diagnosis and all but 1 died within 1 year of diagnosis. More than half of reported cases had renal involvement and three-fourths had liver infiltration of leukaemic cells with around 50% having hypoglycaemia. Moreover, out of seven children in their study, two had T-ALL (both relapse cases) and both were refractory to chemotherapy and died due to progressive disease.¹ The index child tolerated the intensive phase of therapy well, and the resolution of LA was complete and much earlier as compared with other published reports.1 3 10 11 Milder phenotype of WE in index child is probably due to low TLC at presentation with low tumour burden, non-involvement of liver and kidney as evident by normal liver function tests and normal echotexture on ultrasonography and absence of hypoglycaemia. A similar milder presentation of type B LA has been described in a case with leukaemia presenting with low TLC and associated renal tubular acidosis.¹¹

We therefore hypothesise that tumour type, tumour burden, organ involvement and associated mutations which alter metabolic pathway at time of presentation may have implication on the phenotype of type B LA. A detailed genetic study of associated mutations may help us in delineating phenotypes of type B LA.

Patient's perspective.

Mother’s perspective: My daughter was losing weight for 3 months due to poor appetite and recurrent vomiting. I waited for spontaneous resolution of symptoms. But when her respiratory distress worsened, I took a medical consultation. She was subsequently diagnosed to have blood cancer. There was a remarkable improvement in general condition of my child after 2–3 days of start of chemotherapy. She regained her strength and started gaining weight in 2–3 weeks after chemotherapy. Presently she has gained 6 kg of weight and has completed intensive phase of therapy.

Learning points.

• Type B lactic acidosis (LA) should be suspected in children with malignancy presenting with LA without associated organ dysfunction.

• Type B LA should always be kept in differential diagnosis at the time of initial presentation of haematological malignancies with LA and/or hypoglycaemia.

• Management of the underlying malignancy should get precedence in management of type B LA.

• Type B LA associated with haematological malignancy is generally an extremely poor prognostic marker; however, a milder phenotype as described in index child may exist.