A clinicopathological study of eight patients with COVID-19 pneumonia and a late-onset exanthema.

Editor

Our hospital diagnosed the two first non‐imported COVID‐19 cases in Spain on 26 February. Up to date, a total of 1177 COVID‐19‐infected patients have been hospitalized. Eight of them were referred for dermatological examination. Due to the overwhelming situation, the incidence of cutaneous manifestations in our hospitalized COVID‐19‐infected patients is probably greater than this 0.7%. ,

We studied four males and four females with a mean age of 72.2 years (Table 1). The mean hospital stay was 23.2 days. Two patients required intensive care. No patient has died so far.

Table 1

Clinical and histopathological findings in our patients

Patients	Gender/age (years)	Nasopharyngeal swab	Rash distribution	Lesions type	Cutaneous latency time (days)	Duration (days)	New drugs in the previous 2 weeks	Days of evolution when biopsied	Histopathological findings
Case 1	M/58	Positive	Generalized	Coalescent erythematous‐violaceous maculopapules	29	12	None	4	Subcorneal pustules, spongiosis, papillary oedema, dense perivascular and interstitial neutrophilic infiltrate with moderate presence of eosinophils, erythrocyte extravasation, fibrin thrombi, melanophages
Case 2	F/84	Negative	Trunk, flexures	Coalescent erythematous maculopapules	12	11	Hydroxychloroquine, lopinavir/ritonavir, ceftriaxone	2	Subcorneal pustules, spongiosis, papillary oedema, moderate perivascular and interstitial neutrophilic infiltrate with discrete presence of eosinophils, erythrocyte extravasation, focal fibrin thrombi
Case 3	F/82	Positive	Trunk, flexures	Ill‐defined erythematous patches	29	16 and ongoing	Fosfomycin	1	Intraepidermal pustules, spongiosis, discrete perivascular and interstitial neutrophilic infiltrate with scarce presence of eosinophils
Case 4	F/68	Positive	Trunk, flexures	Ill‐defined erythematous patches	28	9	Metamizole, linezolid, piperacillin‐tazobactam, amiodarone	2	Subcorneal pustules, spongiosis, papillary oedema, discrete perivascular and interstitial neutrophilic infiltrate with scarce presence of eosinophils
Case 5	M/51	Positive	Trunk, proximal extremities	Coalescent erythematous macules	29	10	None	5	Focal spongiosis, exocytosis of neutrophils, discrete perivascular and interstitial neutrophilic infiltrate with discrete presence of eosinophils, focal fibrin thrombi, focal basal layer vacuolar degeneration
Case 6	M/88	Positive	Trunk, proximal extremities, face	Coalescent erythematous maculopapules	31	12	Furosemide	1	Subcorneal pustules, spongiosis, presence of necrotic keratinocytes, papillary oedema, discrete perivascular and interstitial neutrophilic infiltrate with scarce presence of eosinophils, melanophages
Case 7	F/69	Positive	Trunk, flexures, face	Coalescent erythematous maculopapules, pustules, desquamation	33	15 and ongoing	None	6	Subcorneal pustules, spongiosis, papillary oedema, moderate perivascular and interstitial neutrophilic infiltrate with discrete presence of eosinophils
Case 8	M/78	Positive	Trunk	Ill‐defined erythematous patches	30	8 and ongoing	Piperacillin‐tazobactam, meropenem, linezolid	4	Spongiosis, discrete perivascular and interstitial neutrophilic infiltrate with scarce presence of eosinophils

F, female; M, male.

Analytically, all the patients presented lymphopenia and elevated D‐dimer and C‐reactive protein. Patient 1 presented also neutrophilia, eosinophilia and elevated liver enzymes.

Most frequent drugs during hospitalization were hydroxychloroquine (HCQ), lopinavir/ritonavir and ceftriaxone. These treatments had finished at least 1 week before the onset of the cutaneous rash in all cases, except patient 2. Three patients had no new medications in the previous 15 days. Only two patients received new drugs in the previous 48 h.

Mean latency time from systemic symptoms to exanthema was 27.6 days. Exanthema varied from ill‐defined erythematous patches to coalescent maculopapules, some of them with violaceous centre (Fig. 1). The trunk was invariably involved with predilection for the back and folds. Patient 7 developed pustules and desquamation (Fig. 1f). Mean duration of the exanthema is 11.6 days so far.

Figure 1

(a) Patient 1 showing erythematous maculopapules, some of them with slight violaceous centre, coalescing on the axillar area. (b) Erythematous maculopapules coalescing on the back and buttocks of patient 2. (c) Widespread ill‐defined erythematous patches on the trunk of patient 4. (d) Ill‐defined coalescent erythematous macules on the back of patient 5. (e) Patient 6 showing erythematous maculopapules coalescing on torso, axillar and inguinal folds. (f) Patient 7 showing intense erythematous maculopapules coalescing on the back. Some areas of non‐follicular pustules and postpustular desquamation can be appreciated.

Histologically, patients showed in a variable degree: spongiotic dermatitis, non‐follicular subcorneal pustules, neutrophilic exocytosis, an interstitial neutrophilic infiltrate and scarce eosinophils (Table 1, Fig. 2). Three patients presented signs of vascular injury with microthrombi inside dermal capillaries and hematic extravasation (Fig. 2a,b). Fibrinoid necrosis of the vessel walls was not found. The presence of microorganisms was ruled out.

Figure 2

(a) Case1 (H&E, 10×). Subcorneal pustule, a dense perivascular and interstitial mixed neutrophilic and lymphocytic infiltrate with moderate presence of eosinophils and erythrocyte extravasation. (b) Case 1 (H&E, 40×). Fibrin thrombi in papillary dermis capillaries and erythrocyte extravasation. Fibrinoid necrosis of the vessel walls was not found. (c) Case 1 (H&E, 40×). More detailed view of the perivascular and interstitial neutrophilic infiltrate and spongiosis. (d) Case 2 (H&E, 40×). Subcorneal pustule and spongiosis.

Cutaneous manifestations of COVID‐19 have recently been classified in five different clinical patterns .

Our patients presented maculopapular eruptions that shared a late onset. This fact leads us to hypothesize that the rash could be caused by the cytokine storm of the hyperinflammatory phase rather than by the virus itself.

A common histological pattern was found, and it was more striking in patients whose biopsies were performed later; therefore, it might be the same process in different stage of evolution. Three patients showed microthrombi in dermal capillaries as the ones described by Gianotti et al. in two cases. Recent publications have found evidence of direct viral infection of the endothelial cell in different organs that could explain the impaired microcirculatory function of this disease . In our cases, microthrombi could be a clue for severe acute respiratory syndrome coronavirus‐2 (SARS‐CoV‐2) infection. They could also be an epiphenomenon of a possible acute generalized exanthematous pustulosis (AGEP), but this has not been described before. A further in situ hybridization or SARS‐CoV‐2 immunohistochemistry would help to clarify this question.

We found difficulty to distinguish the viral or drug origin of the rashes. However, only two patients started new medications during the few days before, as usually occurs in AGEP. Furthermore, clinically no patient except one resembled a classical AGEP. Patch tests need to be performed to determine a possible drug responsibility.

An atypical form of AGEP, generalized pustular figurate erythema, has been described as a delayed reaction to HCQ. Nevertheless, the incidence of this adverse effect to HCQ is infrequent. Thus, if the type of exanthema we are describing here is related to HCQ, there may be another factors (as SARS‐CoV‐2 virus itself) playing a role.

To conclude, we present a series of eight patients with COVID‐19 pneumonia who developed a late‐onset maculopapular exanthema with a distinct histological pattern. Microthrombi may be a clue for SARS‐CoV‐2 endothelial infection.

Acknowledgements

We thank all the health workers of Hospital Universitario de Torrejón for their extraordinary work during this pandemic. All patients in this manuscript have given written informed consent to the publication of their case details and pictures.

Conflict of interest

None.

Funding sources

None.