Erythema and ecchymosis after newborn’s phototherapy

May 1, 2017

A male infant is born and delivery is remarkable for yellow amniotic fluid and a jaundiced infant. Following delivery he is given intensive phototherapy and then develops erythema, which later becomes ecchymosis. What's the diagnosis?

A 2660-g male infant was born via cesarean delivery to a G3P1011 32-year-old mother at 38 weeks 5 days. The mother’s blood type was A-negative with known anti-D (big D) and anti-C (big C) antibodies. She had not received adequate prenatal care and did not get antenatal RhoGAM (Rho [D] immune globulin human). Furthermore, she had known hepatitis C with a high viral load and was a polysubstance abuser. Notably, the infant had a normal middle cerebral artery (MCA) Doppler 18 days prior to delivery. Delivery was remarkable for yellow amniotic fluid and a jaundiced infant. His Apgar scores were 9 and 9 at 1 and 5 minutes. Cord pH was 7.337, and the infant’s blood type showed direct antiglobulin test (DAT)-positive and O-positive. Per protocol, the patient was started on intensive phototherapy and the bilirubin peaked at 17.0 mg/dL at 7 hours of age.

The patient was given a dose of intravenous immunoglobulin G (IVIG) and underwent a double-volume exchange transfusion. He received 2 more doses of IVIG following the transfusion and 2 units of platelets secondary to thrombocytopenia. He remained on intensive phototherapy until day of life 4 when his bilirubin fell below treatment light level at 8 mg/dL, and phototherapy was discontinued. He also developed transaminitis with a peak aspartate transaminase (AST) of 277 IU/L and alanine transaminase (ALT) of 66 IU/L at 34 hours of life. These normalized by day of life 6.

Seven hours following the initiation of phototherapy (at 9 hours of life), the infant began to have erythema of his chest and abdomen to light-exposed areas. Over the course of the next 12 hours, this area became ecchymotic. The area increased in size to cover the majority of the trunk sparing his limbs and head. There was a sharp demarcation where cutaneous cardiorespiratory monitor leads were placed.

Recommended: Recurrent brown spots in an infant girl

The area of erythema and purpura was flat and there was no temperature differential between the area of involvement and the surrounding normal-appearing skin. The infant remained vigorous and was not in distress with palpation of the area. His exam was otherwise unremarkable with no increased work of breathing, normal newborn reflexes, clear breath sounds throughout, and no murmur on the cardiac exam. The wound management team, plastic surgery, and dermatology were consulted on day of life 4 to assist with diagnosis and treatment of this lesion (Figure 1).

Due to increasing WAT-1 (Withdrawal Assessment Tool) scores and known maternal drug use, the infant was placed on the neonatal abstinence syndrome (NAS) withdrawal protocol at 48 hours of life.

Differential diagnosis

The differential diagnosis for this infant’s striking skin findings was broad (Table 1).

Risk management was initially involved given concerns for a burn. One thought was an ultraviolet-mediated burn from the phototherapy spotlight. This seemed unlikely given there was no blistering or sloughing of the skin. Further, on palpation, the lesion was not warm or tender. For similar reasons, a thermal burn was not likely.1

Infectious etiologies were entertained. Given that the mother was group B Streptococcus (GBS) positive and had hepatitis C, the likely inciting diseases were plentiful. Cytomegalovirus (CMV) and enterovirus panels were checked and eventually came back negative. There was suspicion for vesicular lesions on the mother’s labia. Herpes simplex virus (HSV) workup was initiated, including HSV polymerase chain reaction (PCR) from blood and cerebrospinal fluid (CSF), as well as surface cultures, all of which were negative. Acyclovir was initiated and continued for 10 days of therapy. Furthermore, the patient’s lesion did not appear herpetic in nature, as it lacked a clustered vesicular presentation. Septic skin lesions also seemed unlikely because the petechial eruption was confined to a well-defined area, and the infant did not appear ill.

Given the initial thrombocytopenia and transaminitis, a hepatic etiology was considered. However, the thrombocytopenia and elevated liver enzymes resolved quickly. Accidental trauma during the exchange transfusion and thrombocytopenia resulting from the exchange transfusion, as well as vitamin K deficiency, subcutaneous fat necrosis, and unusual vascular malformation, also were considered, but the distribution, sharply demarcated borders, and morphology were not consistent with these diagnoses.

More: Treat plays trick on 3-year-old boy

Inherited disorders of metabolism, namely tyrosinemia, were considered given the bilirubinemia and perceived liver dysfunction. Neonates with tyrosinemia also have a tendency to bruise easily. However, this infant did not have any of the other expected features such as fever, lethargy, vomiting, melena, cabbage-like odor, and failure to gain weight.

Drug reaction also was considered. Initial anti-infectious agents were started shortly after birth. Certain drugs are known to increase light sensitivity, such as furosemide. However, upon review of the patient’s medication regimen, no light-sensitive medications were found.

Finally, a rare hematologic disorder, a congenital variant of porphyria, was considered. These cases usually have severe cutaneous and neurologic manifestations including bullae and blistering.

Finally, a diagnosis

After a literature review and coordination with the medical librarian, concern for transient porphyrinemia was considered. The clinicians collected urine porphyrins, free erythrocyte porphyrin, and porphyrin fractionated plasma, which revealed elevation of all these values. Repeat testing showed a decreasing trend by day of life 22, and no further lesions developed during the infant’s inpatient stay.

Supportive care and monitoring of the patient’s cutaneous eruption was provided. Following the completion of phototherapy on day of life 4, improvement of the lesion was noted (Figure 2) and resolution without treatment was complete 8 days later on day of life 12 (Figure 3). The patient was discharged at 4 weeks of age after completing his morphine wean of the NAS protocol. Repeat urine porphyrins were obtained at 5 months of age and all had nearly normalized (Table 2).2  


Upon literature review, 11 other cases of “transient porphyrinemia” have been reported since the 1980s. No consensus on the mechanism of this phenomenon has been reached. In the authors’ novel side-by-side review of these published cases, several themes occur, such as hemolytic disease of the newborn, the need for transfusions, elevation of porphyrin levels, time to resolution of the purpuric lesion, thrombocytopenia, and transaminitis.3 (See Table 3.) Unlike this patient, many of the previous patients had direct hyperbilirubinemia and exposure to light-sensitizing medications such as furosemide.1 The bilirubinemia may suggest hepatic immaturity.4 Hemolytic disease in the newborn occurs with some frequency at most major neonatology centers, and the authors note that their center sees around 1 case a week. Everything that was done to this patient-using an overhead light, exchange transfusion, and IVIG-has been done to hundreds of other patients without previously witnessing this effect at the authors’ institution. If this is a hereditary entity, factors such as ethnicity and gender would be a consideration; the prior reported cases do not include this information.

There are at least 9 different categories of porphyrias that are classified based on the predominant substrate that is elevated in the serum, urine, or feces. (See Table 3.) The majority of these known enzyme defects are traced to the liver; this is where the synthesized heme (or incomplete substrate) is incorporated into the hemoglobin molecule (Figure 4). Specifically, the last couple of intermediates cause known cutaneous skin changes (ending in -ogen: uroporphyrinogen III, coproporphyrinogen III, protoporphyrinogen).5 The important aspect to note in all these cases is the global elevation of substrates, not suggesting 1 specific category of known porphyria.

A plausible hypothesis would be a reduced functioning of the ultimate enzyme in the pathway-ferrochelatase. Ferrochelatase utilizes iron to bind to the protoporphyrin to form the heme molecule. Perhaps transient deficiency of iron available in the initial period could trigger the reaction. Specific antibodies either from infant or mother could be mediated against the enzyme protein. This could be a genetic variant of this enzyme that has a slower activity when in a stress state (such as the UGT-1A1 in Gilbert syndrome). A heterozygote mutation could give partial function of the enzyme; for example, porphyria cutanea tarda (PCT) often is not manifested in adults because 50% of the enzyme functioning is sufficient.

Regardless of the exact etiology of this phenomenon, it is important to be aware of this developing complication during phototherapy in the neonatal period. Blue wavelengths of light used for phototherapy are known to be important in activation of porphyrins in porphyria. Although the incidence is unknown, it is possible that many neonates display elevated levels of these porphyrin intermediates. However, if they do not require phototherapy, then cutaneous light reaction doesn’t occur.

Next: Girl with rash and muscle weakness

In all but 1 of the reported cases, patients have not displayed any porphyria manifestations later in life. More cases and characteristics of patients are needed to further classify and possibly risk stratify neonates in the future. It appears that even if phototherapy can induce this effect, the benefits of lowering bilirubin levels to prevent kernicterus outweigh the risks of developing this purpuric rash.


1. Paller AS, Eramo LR, Farrell EE, Millard DD, Honig PJ, Cunningham BB. Purpuric phototherapy-induced eruption in transfused neonates: relation to transient porphyrinemia. Pediatrics. 1997;100(3 pt 1):360-364.

2. Ozalla MD, Herrero C, Ventura PJ, Lecha M, Alvarez L, Mascaro JM. Urinary porphyrin excretion measurements in healthy neonates. J Dermatol. 1999;26(7):409-415.

3. LaRusso J, Wilson J, Ceilley R. Phototherapy-induced purpuric eruption in a neonate. J Clin Aesthet Dermatol. 2015;8(3):46-48.

4. Villanueva M, Fulbright JM, Horii KA. Phototherapy induced cutaneous eruption in a neonate with transient porphyrinemia. J Neonatal Perinatal Med. 2015;8(4):417-420.

5. Tuerk MJ, Konia TH, Lamb PM. Part III: Phototherapy-induced purpura in the setting of transient porphyrinemia in a neonate. J Drugs Dermatol. 2011;10(3):306-307.

6. By CC, Owens A, Wesson SK. Purpuric eruption in a transfused neonate receiving phototherapy. Pediatr Dermatol. 2014;31(6):e152-e153.

7. Crawford RI, Lawlor ER, Wadsworth LD, Prendiville JS. Transient erythroporphyria of infancy. J Am Acad Dermatol. 1996;35(5 pt 2):833-834.

8. Mallon E, Wojnarowska F, Hope P, Elder G. Neonatal bullous eruption as a result of transient porphyrinemia in a premature infant with hemolytic disease of the newborn. J Am Acad Dermatol. 1995;33(2 pt 2):333-336.

9. Elumalai J, Spoto-Cannons A, Subramanian N, et al. Index of suspicion in the nursery. NeoReviews. 2010;11(5):e257-e260.

10. Jenkins SM, Cipriano S, Fung, C. Rash associated with phototherapy in a 2-day-old preterm male infant. NeoReviews. 2016;17(1):e55-e59.