218

Original Persistent, Intermitent Delayed Swelling

Articles PIDS: late adverse reaction to Hyaluronic
Acid fillers
Authors:
Fernanda Aquino Cavallieri1 Edema tardio intermitente e persistente ETIP: reação adversa tardia
Laila Klotz de Almeida Balassiano2
Julien Totti de Bastos3 ao preenchedor de ácido hialurônico
Gabriela Helena Munhoz da Fontoura4
Ada Trindade de Almeida5 DOI: http://dx.doi.org/10.5935/scd1984-8773.201793931

1

Radiologist, member of the Com-
mittee of Ultrasound of the Brazilian ABSTRACT
Society of Ultrasound, specialized in Introduction: The use of hyaluronic acid fillers (HA) for esthetic corrections has expo-
Ultrasound by the Brazilian College of nentially grown in recent years. Skin ultrasound (US) is an effective method to identify
Radiology, medical director at Clínica the filler and its complications. A particular type of adverse effect, characterized by late,
Cavallieri, RJ, Brazil.
persistent edema of an intermittent nature, has been lately observed.
2
Dermatologist at Clínica Dermatológi-
ca Dra Gisele D. M. Torok. Specializa- Objective: To describe a delayed complication due to easy HA fillers, named by the au-
tion in Dermatology by Policlínica thors as Persistent, Intermitent Delayed Swelling (PIDS).
Geral do Rio de Janeiro - Rio de Janei- Methods: From October 2016 to July 2017, US examinations performed at Cavallieri
ro (RJ), Brazil. Clinic were selected and referred for evaluation of post-filler complications. Questionnai-
3 
Dermatologist at Instituto Feminino
res were sent to requesting physicians for clinical data collection.
Fetal, Curitiba (PR), Specialist in Der-
matology by Hospital Federal da La- Results: Of 108 exams, 33 cases of local subcutaneous edema associated with the presence
goa - Rio de Janeiro (RJ), Brazil. of HA fillers were identified. Episodes of edema were referred to as recurrent, in the pre-
4
Dermatologist at Clínica Dermatológi- viously affected area, or at another injection site.
ca Dra. Gabriela Munhoz, Rio de Janei- Conclusion: The authors propose a specific nomenclature: PIDS to refer to this late ad-
ro (RJ), Specialist in Dermatology by
verse reaction to HA fillers, which includes delayed and intermittent local swelling, trigge-
Universidade Federal do Rio de Janei-
ro - Rio de Janeiro (RJ), Brasil. red by specific conditions, that persists for as long the HA remains in the subcutaneous
5 
Dermatologist, Preceptor of the Cos- tissue.
miatry Outpatient Clinic, Hospital do Keywords: dermal fillers; hyaluronic acid; adverse effects; skin ultrasound
Servidor Público Municipal de São
Paulo - São Paulo, Brazil

RESU­MO
Introdução: O uso do ácido hialurônico para correções estéticas cresceu exponencialmente nos últimos
anos. O ultrassom de pele mostra-se método eficaz para identificação do preenchedor e suas complicações.
Um tipo particular de efeito adverso, caracterizado por edema tardio e persistente, de caráter intermitente,
Correspondence to:
vem sendo observado ultimamente.
Dra. Fernanda Cavallieri 
Rua Carlos Gois, 375, Sala C307 Objetivo: Caracterizar uma complicação tardia após preenchimento facial com ácido hialurônico.
Leblon, Rio de Janeiro - RJ Brasil. Métodos: Selecionaram-se exames de ultrassom da pele realizados em clínica privada de outubro de
CEP: 22440-040 2016 a julho de 2017, encaminhados para avaliação de complicação após preenchedores. Questionários
E-mail: [email protected] foram enviados aos médicos solicitantes para coleta de dados clínicos.
Resultados: Em 108 exames foram identificados 33 casos de edema local associado à presença de ácido
Received on: 09/11/2017 hialurônico. Episódios de edema foram referidos como recorrentes, na área previamente afetada ou em
Approved on: 16/03/2017 outro sítio de injeção.
Conclusão: Os Authors propõem nomenclatura específica: edema tardio intermitente e persistente para
agrupar as reações adversas tardias ao ácido hialurônico, que se traduzem por edema local tardio, de ca-
Study conducted at Hospital Federal da
ráter intermitente, deflagrado por gatilhos específicos e que persiste enquanto houver a presença do ácido
Lagoa - Rio de Janeiro(RJ), Brasil.
hialurônico no tecido.
Financial support: none Palavras-chave: preenchedores dérmicos; ácido hialurônico; efeitos adversos; ultrassom de pele
Conflict of interests: none

Surg Cosmet Dermatol 2017;9(3):218-22.

PIDS: HA adverse reaction 219

INTRODUCTION
The use of fillers, particularly of hyaluronic acid (HA), in skin and soft tissue ultrasound. The device used was EPIQ7
for facial aesthetic corrections increased exponentially over the (Philips Medical Systems, Bothell,WA, USA), with two high fre-
last few years. Skin ultrasound (US) has been shown to be an quency transducers (7 to 15MHz and 5 to 18 Mhz). The exam
effective method for the evaluation of the substance injected, included the study of the whole face in all patients, in B mode
as well as its complications. The technique is useful because it is and with the association of Color Doppler for the assessment
not invasive, provides a good balance between penetration and of local vasculature. After recording demographic and image
image resolution, enables distinction of the different skin layers data, the authors created a questionnaire that included questions
and presents no risk or discomfort for the patient, nor radiologic about: commercial name of the HA used, application sites, time
exposure, use of contrasts ou confinement in small spaces. for the onset of symptoms, related events, treatment used, recur-
1
According to Ximena et al , the sonographic appearance rence and duration of edema. The questionnaires were sent by
of injected HA has a round or oval-shape, well-defined, anecho- e-mail to the referring doctors for clinical data collection.
ic (black) structure, known as “pseudocyst” because of its re-
semblance with true cysts (Figure 1). HA formulations that are RESULTS
mixed with lidocaine present with interspersed echoes (debris In 108 ultrasound exams, performed for the evaluation
inside the pseudocysts). Polymethylmethacrylate (PMMA) has of complications with fillers, 33 cases of subcutaneous edema
the sonographic aspect of multiple hyperechoic deposits (white), associated to HA were seen, all of them in women between
that cause a mini-artefact shaped like a comet tail, corresponding 29 and 71 years of age, (mean of 50 years). In 27 exams, HA
to the posterior reverberation. Because calcium hydroxyapatite was identified as the only filler; in six patients, besides HA, an-
is a compound of microspheres suspended in a lipopolysaccha- other substance was visualized, namely: polymethylmethacrylate
ride carrier, it is identified by hyperechoic deposits with vari- (PMMA) in three exams, Poly-L-Lactic acid (PLLA) in one
able degrees of acoustic shadow. Silicone oil appears as a strong- exam, autologous fat in one exam and polyacrylamide gel in one
ly echogenic image in the subcutaneous tissue, determining a exam. Regarding the presence of the filler, the most affected site
strong posterior acoustic shadow in a snowstorm pattern. The by edema was the malar region (15 cases), followed by the lower
sonographic aspect of polyacrylamide gel is of an anechoic, oval- eyelid region (11), nasolabial fold (eight) and lip (two). Zygoma,
shaped pseudocyst with hyperechoic lines (white), that does not chin, pre-jowl region, forehead and nose had one case each. Of
change in volume over time and that determines an enhance- the 33 affected patients, five presented with edema in different
ment of the echogenicity of the surrounding tissue. Autologous areas at the same time. Of the 33 patients, the exam was repeated
fat filler is visualized as an oval shaped, well-defined, isoechoic in 12, due to recurrences either in the same site of injection or
(similar to the fat in the adjacent subcutaneous tissue) nodulari- in other sites.
ty, sometimes with minute interspersed anechoic areas. Polylac- Among the sonographic findings, the common feature
tic acid usually has no sonographic expression, except in cases in all 33 cases was the presence of HA associated to a diffuse
where the product acquires a nodular aspect and becomes clin- increase in the thickness and echogenicity of the surrounding
ically palpable, when it is visualized on ultrasound as a well-de- subcutaneous tissue, sonographically similar to a diffuse, ill-de-
fined isoechoic image. fined panniculitis, corresponding to the area with clinical edema
Regarding the complications caused by fillers, US can (Figure 2). No solid nodules or collections of liquid were seen in
identify the filler substance, determine its dimensions and loca- these patients, therefore ruling out other possible adverse reac-
tion, and evaluate local vasculature with color Doppler. Ultra- tions that are characterized by nodules, abscesses or collections.
sound images differ from inflammatory and/or infectious process- These edema episodes were referred as recurrent by the patients,
es, overcorrections and changes consistent with necrosis of the occurring in the area previously affected or in another injec-
subcutaneous tissue. The test can also help in guiding aspiration tion site. By the features of delayed edema after use of HA filler,
1-3
biopsies and hyaluronidase and/or corticosteroid injections. intermittent and persistent in nature, all 33 patients were diag-
Recently, a particular type of complication that evolves nosed as PIDS (persistent intermittent delayed swelling).
with delayed, recurrent and persistent facial edema, correspond- Of these 33 cases of PIDS, data of 20 patients were col-
ing to the injection site, drew the authors’ attention for being the lected through questionnaires. PIDS was characterized clinically
reason for frequent requests for facial soft tissue ultrasound ex- by non-pitting, erythematous or not, diffuse or not, ill or
amination. The intention to better clarify the clinical features of well-defined edemas along the area of HA injection. In all cases,
this type of complication lead the authors to conduct this study. accentuation of the edema was reported upon waking up, with
slight improvement throughout the day.
METHODS The mean duration for each event was difficult to evalu-
In the period from October 2016 to July 2017, all US ate and varied considerably between patients; most of them were
performed at Clínica Cavallieri de Diagnóstico por Imagem, primarily treated with oral corticosteroids and/or antibiotics be-
Rio de Janeiro (RJ), Brazil, referred for the evaluation of com- fore being referred for ultrasound examination. The earliest case
plications of facial fillers were selected. All patients underwent appeared 25 days after the injection, whereas the latest occurred
facial US, performed by a radiologist with a large experience three years after the procedure. Twelve patients had the onset of

Surg Cosmet Dermatol 2017;9(3):218-22.

220 Cavallieri FA, Balassiano LKA, Bastos JT, Fontoura GHM, Almeida AT

Figure 1: To the left, sonographic image of the HA (hyaluronic acid) deposit, interspersed with the subcutaneous tissue in the malar region (between x and
+ markings); to the right: illustration of the sonographic aspect of HA deposits restricted to the subcutaneous tissue

(two cases), oral corticosteroid alone (three cases). One patient

did not receive any treatment and had spontaneous resolution.

DISCUSSION
Clinically recognized, the persistent intermittent delayed
edema consists in recurrent episodes of local edema in the HA
injection site, with short of long remission, with no evidence
of defined palpable nodules. On the ultrasound, the presence
of HA corresponding to the edematous area is seen, associated
to a diffuse increase in the thickening and echogenicity of the
subcutaneous tissue (panniculitis).
Ultrasound is a non-invasive imaging study, of easy ac-
cess, that is being frequently used in dermatological practice.
Ultrasound offers relevant information on adverse reactions of
cosmetic fillers, being an important tool in cosmiatry for the
better understanding of complications post-fillers.
Many HA complications have been described in the lit-
Figure 2: Upper: clinical image of left lower eyelid edema in a site of pre- erature, however, each author classifies them according to their
vious injection of HA (Juvederm Volbela®). To the left: sonographic image
clinical experience because there is no consensus on the classi-
of the presence of HA deposits (between x and +) associated to the increase
in thickness and echogenicity of the subcutaneous tissue surrounding the fication of these adverse reactions. Nonetheless, many articles
left lower eyelid. To the right: sonographic comparison of the right lower describe a late adverse reaction similar to PIDS.
4
eyelid (not affected) with the left lower eyelid of the patient, demonstra- Callan et al reported a single case (1%) of “edema and
ting the increase in thickness and echogenicity of the subcutaneous tissue hardening of the product” in the injected area in a 24-month
study with 103 patients treated with Juvederm Voluma®.
5
Goodman also described a “firm and hardened,
non-pitting edema, with no signs of infection or inflamma-
tion” in a patient treated with Juvederm Voluma®, 4 months
PIDS at the same time as an infectious process (sinusitis, urinary after the procedure.
tract infection, respiratory tract infection, dental infection), trau- In a retrospective review of 4,702 treatments with Juve-
derm Voluma® in 2,342 patients, Beleznay et al described 23
6
ma on the face or vaccination. In one case, the patient noted
recurrences of the facial edema during menses. Of the products cases (1%) of “firm nodules and local edema of late onset”. The
utilized, eight were identified: JuvedermVoluma® (seven cases), time for the development of these nodules was of 4 months,
JuvedermVolbella® (seven cases), Juvederm Volift® (two cases), with mean resolution after 6 weeks.
Juvederm Ultra (one case) Anteis Modelis® (one case), Restylane
7
Artzi et al reported a series of 400 patients injected with
Perlane® (one case) and Emervel Classic® (one case). Medical Juvederm Volbella® in the lips and lower eyelids. Of those, 17
management consisted in systemic antibiotic therapy (ATB) + (4.25%) developed “change in color and edema in the treated
hyaluronidase (two cases), ATB + systemic and/or intralesional area”, with 8 of these 17 patients having an association with
corticosteroid (five cases),ATB + hyaluronidase + oral corticoste- other types of HA besides the Juvederm line.The mean time for
roid (five cases), hyaluronidase alone (two cases), oral ATB alone onset was of 8 weeks.

Surg Cosmet Dermatol 2017;9(3):218-22.

PIDS: HA adverse reaction 221

In the above-mentioned articles, ultrasound was not used weight strands are pro-inflammatory, activating the immune sys-
9
to characterize complications. Perez et al reported the use of US tem. It is possible that between three to five months after the
to evaluate one case of complication with HA filler. The patient injection, when the activation of late inflammatory nodules is
had “indurated, palpable and asymptomatic lesions” in the mar- more frequently observed, there is a more pronounced break-
ionette lines that appeared 4 months after the injection with age of the HA, exposing low molecular weight fragments, that
® ®
Juvederm Voluma and Juvederm Volift . US demonstrated a fo- are pro-inflammatory. Even though products from the Juvederm
cal subcutaneous area with enhanced echogenicity, suggestive of Vycross® line have a higher proportion of low molecular weight
edema, and an increase vascularization of the area (panniculitis). particles, is not known if this proportion alone is more inflam-
The patient progressed with improvement of the lesion, but had matory in comparison to other products.
recurrences in other areas of the face for up to 4 months after Regarding triggering factors, 12 patients (36%) associat-
the first episode. All 33 cases evaluated in the present study re- ed the event to an infectious condition, and this data was close
semble the case in Perez et al article, where HA was identified to what Beleznay described in his article, where 39% of patients
in areas that corresponded to the edema and associated to the reported a respiratory tract infection or a dental procedure be-
6
subcutaneous tissue, with increased thickness and hyperechoge- fore the appearance of the reactions.
8
nicity (signs of panniculitis). Given the spontaneous resolution of the nodules, their
Of the five articles mentioned above, we highlight 43 early onset, short duration and treatment response, including ste-
cases of an edematous reaction after injection of HA fillers. Two roids and hyaluronidase, Beleznay et al defend the idea that these
of them were single cases (Goodman – Juvederm Voluma® and types of reactions seen with HA are more consistent with an
Perez – Juvederm Voluma® and Volift®). Two clinical studies immune-mediated etiology as opposed to biofilm, a commonly
were conducted with a single product and comprise 24 cases implicated mechanism in the literature.The opinion of those au-
(Beleznay – Juvederm Voluma®; Callan – Juvederm Voluma®). thors is that, when HA is injected into a predisposed individual,
Lastly, 17 cases came from private practices (Artzi). Contrary triggers such as respiratory tract infections, dental procedures,
to what we found in the literature, with cases from controlled systemic bacterial or viral infections, vaccination and facial trau-
groups or private practices, the cases analyzed by the authors ma could trigger an inflammatory process corresponding to the
come from different sites, referred by dermatologists from their injected area, given the immunogenic nature of the filler, as well
10
private practices to a radiology clinic, focused on ultrasound. as its capacity of retaining water, configuring the local edema.
Therefore, the statistics were collected at Clínica Cavallieri with In 2013, Alijotas et al selected 235 articles published on
varied patients from Rio de Janeiro metropolitan area, from a PubMed from 2000 to 2012 reporting fillers adverse reactions
total of 30 referring dermatologists. with the aim to report the various types of related adverse events.
In the 20 cases where the performing physician reported The results obtained from this review showed that most of the
the brand of HA, Vycross® line of fillers appeared in a high-
er number of cases (16) compared to other lines of HA fillers
(four). The small number of cases where the filler was identified
does not allow us to conclude a cause/effect relationship with
a specific product line, since this could be the most used brand
of filler. A higher number of cases would be required to reach a
more accurate conclusion.
Vycross® technology is based in the incorporation of
short and long strands of HA to provide a more effective retic-
ulation. Published data suggest that high molecular weight HA
strands are mainly anti-inflammatory, while the low molecular

Figure 4: Patient with two episodes of facial edema in different times af-
ter use of HA filler (Juvederm Volift). Upper: unilateral lower eyelid edema
Figure 3: Left lower eyelid edema after flu vaccination in an area of previou-
along with sinusitis. Lower: right upper lip edema along with urinary tract
sly injected HA (Juvederm Voluma®)
infection

Surg Cosmet Dermatol 2017;9(3):218-22.

222 Cavallieri FA, Balassiano LKA, Bastos JT, Fontoura GHM, Almeida AT

CONCLUSION
late effects are inflammatory or immunomediated in nature, and The authors propose a specific nomenclature: persistent
that factors such as systemic infections could act as a trigger for intermittent delayed edema (PIDS) to group late HA adverse re-
11
these complications. actions, characterized by late local intermittent edema, triggered
The data obtained suggest that PIDS is a manifestation by specific factors, that persists while there is HA in the tissue.
that can occur after the use of HA facial fillers, clinically char- On US, the common finding is the presence of HA with
acterized as: late onset diffuse, non-pitting edema along the area signs of surrounding panniculitis (increased thickness and echo-
of HA injection, (it can appear between week and years after genicity of the subcutaneous tissue, correlating to the clinical as-
HA injection), transient and intermittent and, mainly, persisting pect), and the absence of solid nodules or liquids. Since nodules
while there is HA in the tissue. It is frequently related to some cannot be identified on US, we suggest that PIDS have a spe-
trigger such as local trauma, vaccination (Figure 3) or more cific classification in the group of HA late adverse reactions,
commonly after a local or systemic infectious process such as, for commonly described as a group in the literature. l
example, respiratory infection (Figure 4) or dental procedures,
what could explain its intermittent nature.

DECLARATION OF PARTICIPATION: Julien Totti de Bastos:

Preparation and wording of the manuscript
Fernanda Aquino Cavallieri:
Preparation and wording of the manuscript Gabriela Helena Munhoz de Fontoura:
Preparation and wording of the manuscript
Laila Klotz de Almeida Balassiano:
Preparation and wording of the manuscript Ada Trindade de Almeida:
Supervision and final review of the manuscript

REFERENCES
1. Wortsman X. Common applications of dermatologic sonography. J Ul- 7. Artzi O, Loizides C, Verner I, Landau M. Resistant and recurrent late reac-
trasound Med. 2012; 31(1):97-111. tion to hyaluronic acid-based gel. Dermatol Surg. 2016;42(1):31-7.
2. Wortsman X, Wortsman J, Orlandi C, Cardenas G, Sazunic I, Jemec GB. 8. Pérez-Pérez L, García-Gavín J, Wortsman X, Santos-Briz Á. Delayed Ad-
Ultrasound detection and identification of cosmetic fillers in the skin. J verse Subcutaneous Reaction to a New Family of Hyaluronic Acid Der-
Eur Acad Dermatol Venereol. 2012; 26(3):292-301. mal Fillers With Clinical, Ultrasound, and Histologic Correlation. Derma-
3. Quezada-Gaón N, Wortsman X. Ultrasound-guided hyaluronidase in- tol Surg. 2017;43(4):605-8.
jection in cosmetic complications. J Eur Acad Dermatol Venereol. 2016; 9. Baeva LF, Lyle DB, Rios M, Langone JJ, Lightfoote MM. Different molecu-
30(10):e39-e40. lar weigh hyaluronic acid effects on human macrophage interleukin 1 β
4. Callan P, Goodman GJ, Carlisle I, Liew S, Muzikants P, Scamp T, et al. Ef- production. J Biomed Mater Res A. 2014;102(2):305-14.
ficacy and safety of a hyaluronic acid filler in subjects treated for cor- 10. Luebberding S, Alexiades-Armenakas M. Facial volume augmentation
rection of midface volume deficiency: a 24 month study. Clin Cosmet in 2014: overview of different filler options. J Drugs Dermatol. 2013;
Investig Dermatol. 2013; 6:81-9. 12(12):1339-44.
5. Goodman GJ. An interesting reaction to a high- and low-molecular weight 11. Alijotas-Reig J, Fernández-Figueras MT, Puig L. Inflammatory, imune-
combination hyaluronic acid. Dermatol Surg. 2015; 41(Suppl 1): S164-6. -mediated adverse reactions related to soft tissue dermal fillers. Semin
6. Beleznay K, Carruthers JD, Carruthers A, Mummert ME, Humphrey S. Arthritis Rheum. 2013;43(2):241-58.
Delayed-onset nodules secondary to a smooth cohesive 20 mg/mL
hyaluronic acid filler: cause and management. Dermatol Surg. 2015;
41(8):929-39.

Surg Cosmet Dermatol 2017;9(3):218-22.