Study reference

Study characteristics

Patient characteristics

Index test (I)

Control test (C)

Reference test

Follow-up

Outcome measures

Akita 2021

Type of study:

Phase III, multicenter, single-arm, open-label, clinical trial (HAMAMATSU-ICG study)

Country

Japan

Source of funding:

The present clinical trial was conducted with funding from Hamamatsu Photonics

KK.

Also Supported

In Part by the project “Assessment of the utility of indocyanine green fluorescent lymphography

in diagnosing lymphedema” of the Japan Agency for Medical Research

and Development

Inclusion criteria:

adult patients with secondary lymphedema in the lower or upper extremities who were considered for LVA.

Patients with a diagnosis of clinical stage I to II lymphedema, as defined by the International Society of Lymphology (ISL) were eligible for the present study.

Exclusion criteria:

patients who severely

violate the protocol; and patients whose images are of poor quality and cannot be interpreted (cases of test failure).

N total at baseline (n analysed):

55 patients in treatment phase. 57 patients in assessment phase

Baseline characteristics:

Age: 58,7 +- 10,4, Sex: 100% female. BMI 22,6 +- 2,8. Breast cancer: n = 13 and gynecologic cancer n = 44

ICG lymphography

The technique for ICG fluorescent lymphography was standardized as follows. For the assessment phase and lymphatic vessel identification in the treatment phase, after local anesthesia (0.5% xylocaine), 0.2 mL of 2.5 mg/mL ICG solution (Diagnogreen, 25 mg; Daiichi Sankyo Co, Ltd, Tokyo, Japan) was injected subcutaneously into five sites on the hand or foot of each extremity. The injection sites were determined from the results of previous cadaveric studies that had validated the injection sites as covering as many lymphatic channels as possible. After injection, a near-infrared monitoring camera (Pde-neo; Hamamatsu Photonics KK, Hamamatsu City, Japan) was used for imaging. To evaluate the imaging findings, the body surface was stroked manually to promote the migration of the ICG in the lymphatic vessels, and the examination was completed within 3 hours after injection. To confirm the patency of the anastomotic site in the treatment phase, 0.2 mL of 2.5 mg/mL of ICG was injected subcutaneously into the skin at each anastomotic site after the anastomosis was completed.

None

Lymphoscintigraphy

For lymphoscintigraphy, the limb was judged as normal

and categorized as type 0 when the lymphatic vessel

was depicted linearly in the whole extremity with no

dermal backflow (DBF) pattern, no lymphatic obstruction,

and no additional collateral vessels observed. All other findings,

including a reduced number of stained lymph

nodes, DBF, and lack of tracer migration, were considered

abnormal and categorized into scintigraphy types

I to V.

Length of follow-up:

Not mentioned

Loss-to-follow up:

Not mentioned

The present clinical trial aimed to confirm whether ICG fluorescent

lymphography would be useful in evaluating lymphedema, identifying lymphatic vessels suitable for anastomosis, and

confirming patency of lymphaticovenular anastomosis in patients with secondary lymphedema.

Sensitivity, specificity, PPV, negative predictive value, positive likelihood ratio, negative likelihood ratio, and accuracy.

Study reference

Study characteristics

Patient characteristics

Index test (I)

Control test (C)

Reference test

Follow-up

Outcome measures

Akita 2013

Type of study:

Retrospective comparative study

Country

Japan

Source of funding:

None mentioned

Inclusion criteria:

Patients were divided into a secondary group and an idiopathic group, based on whether

they had secondary lymphoedema after lymph node dissection or were suspected of primary lymphoedema

when other possible causes had been excluded.

Exclusion criteria:

Patients with lymphoedema secondary to filariasis, infection or

trauma were not included in the study.

N total at baseline (n analysed):

A total of 169 limbs in 95 secondary patients and 65 limbs in 39 idiopathic patients

were included in the study.

Baseline characteristics:

The patient population comprised 115 females and 19

males. The average patient age was 58,5 years, ranging  from 9 to 82 years.

ICG lymphography

0,3 ml of ICG was injected subcutaneously into the bilateral first web spaces of the feet. One hour after injection, circumferential fluorescent images of lymphatic drainage channels were obtained using a photodynamic eye infrared camera system (Hamamatsu Photonics K.K., Hamamatsu, Japan). Yamamoto et al. categorised indocyanine green lymphography findings into a normal linear pattern and three abnormal dermal backflow patterns. Typically, the findings change from a normal linear pattern to a splash pattern, followed by a stardust pattern and a diffuse pattern as the severity of lymphoedema increases. In this study, ICG lymphography was considered positive for lymphoedema when the dermal backflow pattern, absent or decreased inguinal/axillary lymph nodes or the absence of ICG dye migration was recorded at 60 and 120 min after injection. In this study, indocyanine green lymphography was classified based on the Yamamoto et al. categories.

None

Lymphoscintigraphy

Lymphoscintigraphy was performed by subcutaneous

injection of a small amount (1,0 ml, 200 MBq) of

technetium-99m-labelled human serum albumin. Injections

were given bilaterally and simultaneously, in the first web

spaces of the feet or hands.

Length of follow-up:

Not mentioned

Loss-to-follow up:

Not mentioned

In secondary oedema, the sensitivity of indocyanine green lymphography, compared

with lymphoscintigraphy, was 0,972, the specificity was 0,548 and the accuracy was 0,816.

When patients with lymphoscintigraphy type I and indocyanine green lymphography stage I were regarded as negative, the sensitivity of the indocyanine green lymphography was 0,978, the specificity was 0,925 and the accuracy was 0,953.

There was a significant positive correlation between the lymphoscintigraphy type and the indocyanine green lymphography

stage.

In idiopathic oedema, the sensitivity of indocyanine green lymphography was 0,974, the specificity was 0,778 and the accuracy was 0,892.

Study reference

Study characteristics

Patient characteristics

Index test (I)

Control test (C)

Reference test

Follow-up

Outcome measures

Hara 2018

Type of study:

Prospective study

Country

Japan

Source of funding:

Not mentioned

*On the same day,

they started wearing a weak pressure compression stocking (tg soft, Lohmann & Rauscher GmbH & Co. KG, Rengsdorf, Germany) or a standard elastic garment.

Inclusion criteria:

Patients who underwent LVA in the Department of Lymphatic and Reconstructive Surgery, Saiseikai Kawaguchi General Hospital, between June and November 2016 were recruited as the multi-injection group, while patients who underwent

LVA between October 2015 and March 2016 were recruited as the control group.

Exclusion criteria:

Patients who underwent LVA between April and

May 2016 were not included because this period was a transitional period. Patients who did not undergo ICG lymphography because of allergy to iodine, patients whose preoperative or postoperative circumference

data was lacking, patients who underwent other surgical procedures simultaneously for the affected limb were also excluded. Additionally, limbs that underwent multi-injection of ICG in the control group were excluded.

N total at baseline (n analysed):

We performed LVA in 77 patients (149 limbs) between June and

November 2016, and 67 patients (130 limbs) between October 2015

and March 2016. The excluded patients are listed in Table 1. Finally,

60 patients (115 limbs) in the multi-injection group and 49 patients

(81 limbs) in the control group brought the patients included in this

study to a total of 109 patients (196 limbs).

Baseline characteristics:

The average

age was 60,8 years old (range: 30–88 years old). The number of

male patients was 4 and the remaining 105 patients were female.

ICG lymphography

We injected 0,1 mL of ICG (0,5% Diagnogreen, Daiichi Pharmaceutical,

Tokyo, Japan) into the first web spaces of the feet or the second

web space of the hands, subcutaneously. We observed the

lymphogram with an infrared camera (Photodynamic Eye; Hamamatsu Photonics, Hamamatsu, Japan).

None

Lymphoscintigraphy

Preoperatively, we performed lymphoscintigraphy in all patients.

We injected 0,3 mL (222 MBq) of 99mTc-labeled albumin radioisotope

into the first web space of each foot or the second web space of the

hand. In patients with lower limb lymphedema, scintigram images

were taken at 15, 30, 60, and 90 minutes after the injection. In

patients with upper limb lymphedema, scintigram images were taken

at 10, 15, 30, and 60 minutes after the injection. Scintigrams were

classified according to the Maegawa classification (International Society

of Lymphology, 2013).

Length of follow-up:

Not mentioned

Loss-to-follow-up:

Not mentioned

The average circumference

change was −1,83% in multi-injection group and −0,34% in control group (P = 0,021). The percentage

of the Ectasis type lymphatic vessels was 59,0% in multi-injection group and 40,2% in

control group (p = 3,30 × 10−5).

Conclusions: By injecting ICG into multiple.

Study reference

Study characteristics

Patient characteristics

Index test (I)

Control test (C)

Reference test

Follow-up

Outcome measures

Mihara 2013

Type of study:

Prospective comparative study

Country

Japan

Source of funding:

Supported by a grant from the Japanese funding program for next generation

world-leading research.

Inclusion criteria:

29 patients with lymphedema (58 limbs, including healthy ones) after gynecologic cancer care who underwent lymphedema treatment at The University of Tokyo and Saiseikai Kawaguchi General Hospital between April 2011 and December 2011.

Exclusion criteria:

Patients whose subjective symptoms persisted for less than 6 months were excluded to rule out indefinite

complaint.

N total at baseline (n analysed):

29 patients

Baseline characteristics: All female; age range, 32-79 years

ICG lymphography

ICG (0,2 mL, Diagnogreen 0,5%; Daiichi Pharmaceutical, Tokyo, Japan) was injected intracutaneously into the bilateral first interdigits. Imaging was performed using a PDE system (Hamamatsu Photonics, Hamamatsu, Japan). Vivid dynamic images of superficial lymphatic flow were obtained within a few minutes after the injection. In this method, the skin surface is illuminated with near-infrared light, and light emitted by ICG absorbed in lymph vessels is imaged using a charge coupled device camera to detect lymph flow in subcutaneous lymph vessels. For staging of lymphedema based on ICG lymphography, the classification proposed by Yamamoto et al was used: normal lymphatic pattern (stage 0; Video A, available online only); abnormal findings, such as splash and stardust patterns, in the inguinal

region only (stage 1); a stardust pattern above the knee, but normal lymph flow below the knee (stage 2); a stardust pattern in the inguinal region over the ankle, but normal lymph flow on the distal side from the ankle (stage 3); a stardust pattern spread to the toes (stage 4); and a diffuse pattern (stage 5; Video B, available online only).

None

Lymphoscintigraphy

A Tc-99m-labeled nanocolloid

was injected subcutaneously (w20 MBq in 0,1 mL per

limb) into the dorsum of the first web space of the foot using a 1-mL syringe and a 27-gauge needle. Penetration of the needle into a blood vessel was excluded by careful aspiration before injection. Massage, warm compresses, and lower limb elevation were not permitted after injection. Half-body anterior and posterior images from the upper abdomen to the toes were obtained with a large field-ofview gamma camera, immediately and at 15, 30, 60, and

90 minutes after tracer injection. Staging based on lymphoscintigraphy

was performed using the classification

proposed by Maekawa et al14: no abnormal findings (type 1), a dermal backflow pattern (DB) above the knee (type 2), DB from above to below the knee (type 3), DB only

below the knee (type 4), and no enhancement (type 5).

Length of follow-up:

Not mentioned

Loss-to-follow up:

Not mentioned

The objective of the study was to compare the utility of lymphoscintigraphy and ICG lymphography for diagnosis of cancer-related secondary lymphedema.

The sensitivity and specificity of lymphoscintigraphy and ICG lymphography were calculated for all limbs and for diagnosis of early lymphedema in affected limbs.

Study reference

Study characteristics

Patient characteristics

Index test (I)

Control test (C)

Reference test

Follow-up

Outcome measures

Yoon 2020

Secondary lymphedema after breast cancer.

Type of study:

Retrospective study

Country:

Korea

Source of funding:

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) No. 2019R1F1A1059801.

Inclusion criteria:

Patients who underwent both lymphoscintigraphy and ICG lymphography for the evaluation of secondary unilateral lymphedema after breast cancer surgery.

Exclusion criteria: Patients who had previous primary lymphedema, history of trauma, metastasis or infection of both arms and patients in whom staging was impossible because of poor image quality or atypical findings.

N total at baseline (n analysed):

A total of 47 patients were recruited for this study.

Baseline characteristics:

All patients were female and their mean age was 55,85 years ±10,51 SD.

ICG lymphography

ICG lymphography images were taken by using a near- infrared camera (Moment K, Gils, Seoul, Korea) to detect 760 nm light-emitting diode and a filter-cutting light below 820 nm. A total of 0,2 mL of ICG (Diagnogreen 2,5 mg/mL) was injected subcutaneously into the bilateral upper extremities at the first and third web space of the hand.

None

Lymphoscintigraphy

All lymphoscintigraphies were obtained as planar images by using a dual-head gamma camera (Infinia; GE Medical Systems, Milwaukee, WI, USA) equipped with low-energy, high- resolution collimators, at a rate of 300 s/frame, with a matrix size of 256 pixels ×256 pixels and a zoom factor of 1,0. The energy discriminator was centered on the 140-keV pho- topeak of Tc-99 m by using a symmetric 20% window.

Length of follow-up:

Not mentioned

Loss-to-follow-up: 

A total of 47 patients were recruited for this study. No patient was excluded from the analysis because of poor imaging quality in the final test, except for 3 patients suspected of focal infection.

The mean  lymphoscintigraphy severity stage, anterior aspect of ADB stage, posterior aspect of ADB stage and modified MDACC stage was 3,77 ±1,66; 3,01 ±1,26; 3,94 ±1,49; and 3,30 ±1,38, respectively.

Study reference

Study characteristics

Patient characteristics

Index test (I)

Control test (C)

Reference test

Follow-up

Outcome measures and effect size

Comments

Abbaci 2019

A. Noguchi 2010

B. Noguchi 2012

C. Ikeda 2012

D. Ikeda 2014

E. Sakurai 2014

F. Noguchi 2016

G. Foster 2018

H. Yamamoto 2011

I. Mihara 2012

J. Yamamoto 2014

K. Akita 2016

L. Tashiro 2017

M. Zaleska 2018

N. Zaleska 2018

O. Qin 2018

P. Chang 2013

Q. Chen 2016

R. Gennaro 2016

S. Gentileschi 2017

T. Winters 2017

U. Engel 2018

V. Akita 2018

W. Visconti 2018

X. Seki 2019

Y. Khan 2019

Type of study:

Systematic review

Search date:

A literature review

was conducted until December 2018.

Number of included studies:

N = 33

Country

SR performed in France

Source of funding:

The authors thank the Institute National du Cancer (INCa) for funding (grant PHRC-K15-222) the ARMoNIC study

(NCT02994225).

Inclusion criteria:

(1)

use of fluorescence ICG imaging, (2) treatment of breast cancer, (3)

breast cancer related lymphedema, (4) in vivo data acquisition on

human tissues, (5) article in English.

Exclusion criteria:

(1) ICG-SLN studies, (2)

reviews, letters and comments on articles, (3) abstracts and

proceedings.

N total at baseline (n analysed):

P. 65

Q. 21

R. 40

S. 16

T. 29

U. 124

V. 189

W. 37

X. 30

Y. 27

Age:

Not reported

A. ICG:  0,1 ml (0,25 mg) inner side of the wrist

B. ICG:  0,1 ml (0,25 mg) inner side of the wrist

C. ICG: 0,5e1ml (5mg) at the upper inner ipsilateral arm

D. ICG:  0,5e1ml (5mg) at the upper inner ipsilateral arm

E. ICG:  0,15 ml subcutaneously into the

interdigital area

F. ICG:  0,1 ml (0,25 mg) inner side of the wrist

G. ICG: 0,5e3 mg/ml medial aspect of the

proximal arm at the intermuscular groove

H. ICG:  0,1 ml ICG 0,25% injected subcutaneously into the bilateral upper extremities at the second web space of the hand and the ulnar border of the palmaris longus tendon at the level of the wrist

I. ICG:  0,2 ml ICG 0,5% injected intracutaneously into the bilateral second interdigits

J. ICG:  0,1 ml 0,25% ICG injected subcutaneously into the second web space of the hand

K. ICG:  0,3 ml ICG injected subcutaneously into the first web space of the dorsal aspect of the hand

L. ICG:  0,2 ml ICG 0,5% intracutaneously injected into the second interdigits of the hand and the ulnar border of the palmaris longus tendon at the level of the wrist

M. ICG:  0,2 ml 0,5% ICG injected subcutaneously into the second and fourth interdigital spaces of the hand

N. ICG: 0,2 ml 0,5% ICG injected subcutaneously into the second

and fourth interdigital spaces of the hand

O. ICG: 0,1 ml ICG 0,25% injected into the distal arm at three separate sites: two interdigital injections and one injection in the volar wrist

P. ICG: 0,01-0,02 ml ICG into each finger

Q. ICG: 0,1 mL 0,25% ICG intradermal into the second and third web spaces

of the hand

R.  ICG: 0,1 mL ICG subcutaneously at the second web space of the hand and the ulnar border of the palmaris longus tendon at the level of the

wrist.

S.  ICG into the second and fourth web spaces of both hands

T. ICG: Not mentioned

U. ICG into the second and fourth web spaces of both hands

V. ICG: Not mentioned

W. ICG:  0,2 mL ICG 5 mg/ml intradermal

X. ICG: 0,1 ml ICG 0,25% intradermal into the second web space of both

hands, at the anterior border of the styloid process of both radii,

and at the anterior border of the styloid process of both ulna

Y. ICG:  0,1-0,2ml of ICG 5 mg/ml injected in the second web space of the

hand and the ulnar border of the hand (both hands)

None

Not clearly mentioned for all studies.

Length of follow-up:

Not mentioned

Loss-to-follow-up:

Not mentioned

ICG lymphography is reported for lymphedema diagnosis, with high

sensitivity and specificity, staging, intraoperative mapping and patency control in lymphaticovenular

anastomosis.

Mihara 2012:  The sensitivity was 1 in ICG lymphography and MRI, 0,62 in lymphoscintigraphy, and 0,33 in CT. These results show that MRI and ICG lymphography are superior to lymphoscintigraphy or CT for diagnosis of lymphedema. In some cases, biopsy findings suggested abnormalities in skin and lymphatic vessels for which lymphoscintigraphy showed no abnormal findings. ICG lymphography showed a dermal backflow pattern in these cases.

Diagnosis and staging of lymphedema are usually based on

patients’ reported experiences (for example, pain and heaviness),

characteristic physical findings (for example, mobility, arm volume and arm circumference) and standard lymphoscintigraphy with the injection of radioisotope-labelled colloids into the interdigital

space to monitor the lymphatic flow along the limb using a gamma

camera.

Lymphoscintigraphy is the gold standard but it is time consuming

and expensive, requiring licensing and a nuclear medicine

department. In addition, its poor spatial resolution makes detailed analysis of the lymphatic ducts impossible

Artikel

Reden van exclusie

Chang 2013

Geïncludeerd in SR Abbaci 2019

Chen 2016

Geïncludeerd in SR Abbaci 2019

Chen 2019

Studie in Chinese taal

Garza 2019

Geen referentie test gebruikt

Gentili 2021

Studie in Italiaanse taal

Hara 2013

Klein patiënten aantal (n = 25)

Hayashi 2013

Klein patiënten aantal (n = 18)

Hayashi 2017

Doel studie, geen referentietest gebruikt

Jorgensen 2021

Geen referentietest gebruikt

Mangialardi 2019

Klein patiënten aantal (n = 20)

Mangialardi 2020

Klein patiënten aantal (n = 20)

Mediana-Rodrigue 2020

Klein patiënten aantal (n = 19)

Mihara 2014

Geen duidelijke referentietest gebruikt

Pons 2019

Doel studie, geen referentietest gebruikt

Qin 2018

Geïncludeerd in SR Abbaci 2019

Rebecca 2018

Klein patiënten aantal (n = 15)

Rodriguez 2020

Doel studie, geen diagnostische accuratesse

Suzuki 2020

Klein patiënten aantal (n = 15)

Tashiro 2015

Onduidelijk of een referentietest werd gebruikt

Unno 2010

Klein patiënten aantal (n = 27)

Wiser 2020

Doel studie

Yamamoto 2011

Geen referentietest gebruikt

Yamamoto 2011

Klein patiënten aantal (n = 28), geen referentietest gebruikt

Yamamoto 2013

Geen referentietest gebruikt

Yamamoto 2015

Geen referentietest gebruikt

Yasanaga 2021

Klein patiënten aantal (n = 18)

Yuan 2019

Doel studie

Yuan 2021

Doel studie

Study:

Akita 2021

A. Risk of Bias

Yes/ no/ unclear

Notes

B. Concern of applicability

Low/ high/ unclear

Notes

1. Patient selection

Was a consecutive or random sample of patients enrolled?

No

Adult patients with secondary

lymphedema in the lower or upper extremities who

were considered for LVA.

Is there concern that the included patients do not match the review question?

Low

Review question and included patients both had secondary lymphedema

Was a case-control design avoided?

Yes

Single-arm study

Did the study avoid inappropriate exclusions?

No

Risk of Bias: High

2. Index test

Were the index test results interpreted without

knowledge of the results of the reference standard?

No

A central judgment committee,

composed of physicians who specialized in the

diagnosis and treatment of lymphedema at institutions

other than the study sites, performed blinded evaluations

of the patient information and imaging studies.

Is there concern that the index test, its conduct, or interpretation differ from the review question?

High

* Because it was difficult for anyone other

than the surgeon who performed the operation to determine

patency, the plastic surgeon in charge of the

research at each institution made this determination.

If a threshold was used, was it pre-specified?

Yes

The ICG fluorescent lymphography findings

were judged as normal when the lymphatic vessel was depicted linearly in the whole extremity, and no DBF patterns, including splash patterns (ie, a meandering image

showing dilated lymphatic vessels due to stasis), were

observed. All other findings, including stage I to V ICG DBF pattern classification, were regarded as abnormal.

Risk of Bias: Low

3. Reference standard

Is the reference standard likely to correctly classify the target condition?

Yes

For lymphoscintigraphy, the limb was judged as normal

and categorized as type 0 when the lymphatic vessel

was depicted linearly in the whole extremity with no

dermal backflow (DBF) pattern, no lymphatic obstruction,

and no additional collateral vessels observed.

Is there concern that the target condition as defined by the reference standard does not match the review question?

Low

Were the reference standard results interpreted without knowledge of the results of the index test?

Yes

A central judgment committee,

composed of physicians who specialized in the

diagnosis and treatment of lymphedema at institutions

other than the study sites, performed blinded evaluations

of the patient information and imaging studies.

Risk of Bias: Low

4. Flow and timing

Was there an appropriate interval between index test(s) and reference standard?

Unclear

N.a.

N.a.

N.a.

Did all patients receive a reference standard?

Yes

Did patients receive the same reference standard?

Yes

Were all patients included in the analysis?

No

N=2 excluded, 1 had normal findings and one patient did not receive an intervention

Risk of Bias: Unclear

Study:

Akita 2013

A. Risk of Bias

Yes/ no/ unclear

Notes

B. Concern of applicability

Low/ high/ unclear

Notes

1. Patient selection

Was a consecutive or random sample of patients enrolled?

No

Retrospective study

Patients with suspected extremity lymphoedema and those in whom treatment of lymphoedema was under consideration were compared

Is there concern that the included patients do not match the review question?

Unclear

Demographic features and co-morbidities are unknown.

Was a case-control design avoided?

No

Patients with suspected lymphoedema were included

Did the study avoid inappropriate exclusions?

Unclear

Patients

with lymphoedema secondary to filariasis, infection or

trauma were not included in the study.

Exclusion of patients with “red flags” for the target condition, who may be

easier to diagnose, may lead to underestimation of diagnostic accuracy.

Risk of Bias: Unclear

2. Index test

Were the index test results interpreted without

knowledge of the results of the reference standard?

Unclear

Interpretation is not mentioned in this study, neither is the order of testing

Is there concern that the index test, its conduct, or interpretation differ from the review question?

Unclear

Interpretation is not mentioned in this study. Classification is (Yamamoto et al.)

If a threshold was used, was it pre-specified?

Yes

Indocyanine green lymphography findings are

categorized into 4 patterns.

Typically, the findings change from a normal linear pattern to a splash pattern, followed by a stardust pattern and a diffuse pattern as the severity of lymphoedema increases.

In this study, indocyanine green lymphography was considered positive for

lymphoedema when the dermal backflow pattern, absent or decreased inguinal/axillary lymph nodes or the absence of indocyanine green dye migration was recorded at 60 and 120 min after injection.

Risk of Bias: Unclear

3. Reference standard

Is the reference standard likely to correctly classify the target condition?

No

No consensus exists regarding the criteria for interpretation of lymphoscintigraphy results.

Using the assumption that

the sensitivity and specificity of lymphoscintigraphy is 1.00.

Is there concern that the target condition as defined by the reference standard does not match the review question?

High

No consensus exists regarding the criteria for interpretation

of lymphoscintigraphy results. Although several

attempts at improving objective interpretation through

quantification of lymphatic function have been published,

no general consensus has been reached to date.

Were the reference standard results interpreted without knowledge of the results of the index test?

Unclear

Interpretation is not mentioned in this study, neither is the order of testing

Risk of Bias: Low

4. Flow and timing

Was there an appropriate interval between index test(s) and reference standard?

Unclear

Not mentioned

N.a.

n.a.

n.a.

Did all patients receive a reference standard?

Yes

Did patients receive the same reference standard?

Yes

Were all patients included in the analysis?

Unclear

Not mentioned

Risk of Bias: Unclear

Study:

Hara 2018

A. Risk of Bias

Yes/ no/ unclear

Notes

B. Concern of applicability

Low/ high/ unclear

Notes

1. Patient selection

Was a consecutive or random sample of patients enrolled?

Yes

Patients who underwent LVA in the

Department of Lymphatic and Reconstructive Surgery, Saiseikai Kawaguchi General Hospital, between June and November 2016

were recruited as the multi-injection group, while patients who underwent

LVA between October 2015 and March 2016 were recruited as

the control group.

Is there concern that the included patients do not match the review question?

Low

The purpose of this prospective study was to elucidate

the efficacy of multi-area injection in indocyanine green (ICG) lymphography in LVA.

Was a case-control design avoided?

No

Sixty patients (115 limbs) in multi-injection group and 49 patients (81 limbs) in control group were included.

Did the study avoid inappropriate exclusions?

Yes

Risk of Bias: High

2. Index test

Were the index test results interpreted without

knowledge of the results of the reference standard?

Unclear

It is not mentioned in this study who interpretated the results of the index test and it this person knew the results of the reference test.

Is there concern that the index test, its conduct, or interpretation differ from the review question?

High

On the contrary, we injected ICG at only one site if we could find

good linear pattern both in lower leg and the thigh, or both in the

forearm and the upper arm with the first injection. Therefore, some of

the patients in this group had only one injection.

If a threshold was used, was it pre-specified?

Yes

Risk of Bias: Unclear

3. Reference standard

Is the reference standard likely to correctly classify the target condition?

No

Lymphoscintigraphy, which is the gold standard method for

examining lymphedema, is not adequate as a preoperative examination

of LVA because of its roughness.

Is there concern that the target condition as defined by the reference standard does not match the review question?

Low

Preoperatively, we performed lymphoscintigraphy in all patients.

We injected 0.3 mL (222 MBq) of 99mTc-labeled albumin radioisotope

into the first web space of each foot or the second web space of the

hand. In patients with lower limb lymphedema, scintigram images

were taken at 15, 30, 60, and 90 minutes after the injection. In

patients with upper limb lymphedema, scintigram images were taken

at 10, 15, 30, and 60 minutes after the injection. Scintigrams were

classified according to the Maegawa classification (International Society

of Lymphology, 2013).

Were the reference standard results interpreted without knowledge of the results of the index test?

Unclear

It is not mentioned in this study who interpretated the results of the index test and it this person knew the results of the reference test.

Risk of Bias: High

4. Flow and timing

Was there an appropriate interval between index test(s) and reference standard?

Unclear

Not mentioned

N.a.

n.a.

n.a.

Did all patients receive a reference standard?

Yes

Did patients receive the same reference standard?

Yes

All patients received lymphoscintigraphy, but there was a difference in ICG lymphography

Were all patients included in the analysis?

Unclear

Not mentioned

Risk of Bias: Unclear

Study:

Mihara 2013

A. Risk of Bias

Yes/ no/ unclear

Notes

B. Concern of applicability

Low/ high/ unclear

Notes

1. Patient selection

Was a consecutive or random sample of patients enrolled?

No

Is there concern that the included patients do not match the review question?

Unclear

Early lymphedema was mentioned in the review question, but the exclusion criterium were subjective symptoms related to lymphedema less than 6 months. It is unclear how long the lymphoedema really existed and the duration of lymphedema varied widely between patients

Was a case-control design avoided?

Unclear

Did the study avoid inappropriate exclusions?

No

Patients whose subjective symptoms persisted for less than 6 months were excluded to rule out indefinite complaint.

Risk of Bias: Unclear

2. Index test

Were the index test results interpreted without

knowledge of the results of the reference standard?

Unclear

Not mentioned

Is there concern that the index test, its conduct, or interpretation differ from the review question?

Low

If a threshold was used, was it pre-specified?

Yes

Risk of Bias: Unclear

3. Reference standard

Is the reference standard likely to correctly classify the target condition?

Yes

Is there concern that the target condition as defined by the reference standard does not match the review question?

Low

Were the reference standard results interpreted without knowledge of the results of the index test?

Unclear

Not mentioned *ICG lymphography was performed within 3 months after lymphoscintigraphy.

Risk of Bias: Unclear

4. Flow and timing

Was there an appropriate interval between index test(s) and reference standard?

No

Ideally results of the index test and reference standard are collected on the same patients at the same time. If there is a delay or if treatment is started between index test and reference standard, misclassification may occur due to recovery or deterioration of the condition.

ICG was performed within 3 months.

N.a.

N.a.

N.a.

Did all patients receive a reference standard?

Yes

All subjects were referred to our department for lower extremity lymphoscintigraphy and

ICG lymphography.

Did patients receive the same reference standard?

Yes

Were all patients included in the analysis?

Yes

All procedures were completed without problems in all patients.

Risk of Bias: Low

Study:

Yoon 2020

A. Risk of Bias

Yes/ no/ unclear

Notes

B. Concern of applicability

Low/ high/ unclear

Notes

1. Patient selection

Was a consecutive or random sample of patients enrolled?

No

Retrospective study

Is there concern that the included patients do not match the review question?

Unclear

Demographic features and co-morbidities are unknown.

Was a case-control design avoided?

Yes

Did the study avoid inappropriate exclusions?

Unclear

Exclusion criteria were patients who had previous primary lymphedema, history of trauma, metastasis or infection of both arms and patients in whom staging was impossible because of poor image quality or atypical findings.

Is having atypical findings making lymphedema ‘difficult to diagnose’’?

Risk of Bias: Unclear

2. Index test

Were the index test results interpreted without

knowledge of the results of the reference standard?

Yes

The blind ICG lymphography staging was conducted without having information on the clinical state and lymphoscintigraphy stage to exclude potential confounding variables.

Is there concern that the index test, its conduct, or interpretation differ from the review question?

Low

If a threshold was used, was it pre-specified?

Yes

The images were classified into Type I - V according to arm dermal backflow (ADB) stage 14 and modified MD Anderson Cancer Center (MDACC) stage 15 based on the dermal back flow pattern and how it was extended.

Risk of Bias: Low

3. Reference standard

Is the reference standard likely to correctly classify the target condition?

Unclear

Of the various lymphoscintigraphy staging criteria introduced so far, a gold standard has not been established, and scoring is performed according to findings that are focused on each stage.

Is there concern that the target condition as defined by the reference standard does not match the review question?

Unclear

See notes first question ROB.

Were the reference standard results interpreted without knowledge of the results of the index test?

Unclear

Not mentioned

Risk of Bias: Unclear

4. Flow and timing

Was there an appropriate interval between index test(s) and reference standard?

Yes

The limitation of this study was that tests were not performed on the same day.

However, the two-week interval between the two tests is thought to have been too short to cause exacerbated lymphatic function that would negatively affect the result.

N.a.

n.a.

n.a.

Did all patients receive a reference standard?

Yes

Did patients receive the same reference standard?

Yes

Were all patients included in the analysis?

No

A total of 47 patients were recruited for this study. No patient was excluded from the analysis because of poor imaging quality in the final test, except for 3 patients suspected of focal infection.

Risk of Bias: Low

Study reference

(first author, publication year)

Appropriate and clearly focused question? ¹

Yes/no

Comprehensive and systematic literature search?²

Yes/partial yes/ no

Description of included and excluded studies?³

Yes/partial yes/ no

Description of relevant characteristics of included studies?⁴

Yes/partial yes/no

Assessment of scientific quality of included studies?⁵

Yes/partial yes/no

Enough similarities between studies to make combining them reasonable?⁶

Yes/partial yes/ no

Potential risk of publication bias taken into account?⁷

Yes/partial yes/ no

Potential conflicts of interest reported?⁸

Yes/partial yes/ no

Abbaci

2019

Unclear, no PICO construct, inclusion criteria are predefined 

Partial yes, multiple database, but no mesh terms and grey literature

No, potential relevant studies are not mentioned

Partial yes, confounders not mentioned

Yes, all clinical studies were assigned a level of evidence (LOE) adapted from the Oxford Centre for Evidence-Based Medicine, on methodology and study design, and from the Centre for Reviews and Dissemination of the University of York.

No, no assessment of heterogeneity, publication bias or any other statistical assessment was planned.

No, no assessment of heterogeneity, publication bias or any other statistical assessment was planned.

Partial yes, conflict of interest from authors is mentioned and support is mentioned, but not for each of the included studies

Studie

Uitvoering ICG lymfografie

Akita et al. 2013

0.3 ml of ICG was injected subcutaneously into the bilateral first web spaces of the feet. One hour after injection, circumferential fluorescent images of lymphatic drainage channels were obtained using a photodynamic eye infrared camera system. Yamamoto et al. categorised indocyanine green lymphography findings into a normal linear pattern and three abnormal dermal backflow patterns. Typically, the findings change from a normal linear pattern to a splash pattern, followed by a stardust pattern and a diffuse pattern as the severity of lymphoedema increases. In this study, ICG lymphography was considered positive for lymphoedema when the dermal backflow pattern, absent or decreased inguinal/axillary lymph nodes or the absence of ICG dye migration was recorded at 60 and 120 min after injection. In this study, indocyanine green lymphography was classified based on the Yamamoto et al. categories.

Akita et al. 2021

The technique for ICG fluorescent lymphography was standardized as follows. For the assessment phase and lymphatic vessel identification in the treatment phase, after local anesthesia (0.5% xylocaine), 0.2 mL of 2.5 mg/mL ICG solution was injected subcutaneously into five sites on the hand or foot of each extremity. The injection sites were determined from the results of previous cadaveric studies that had validated the injection sites as covering as many lymphatic channels as possible. After injection, a near-infrared monitoring camera was used for imaging. To confirm the patency of the anastomotic site in the treatment phase, 0.2 mL of 2.5 mg/mL of ICG was injected subcutaneously into the skin at each anastomotic site after the anastomosis was completed. The ICG fluorescent lymphography findings were judged as normal when the lymphatic vessel was depicted linearly in the whole extremity, and no DBF patterns, including splash patterns (i.e., a meandering image showing dilated lymphatic vessels due to stasis), were observed. All other findings, including stage I to V ICG DBF pattern classification, were regarded as abnormal.

Mihara et al. 2013

ICG (0.2 mL, Diagnogreen 0.5%; was injected intracutaneously into the bilateral first interdigits. Imaging was performed using a PDE system. Vivid dynamic images of superficial lymphatic flow were obtained within a few minutes after the injection. In this method, the skin surface is illuminated with near-infrared light, and light emitted by ICG absorbed in lymph vessels is imaged using a charge coupled device camera to detect lymph flow in subcutaneous lymph vessels. For staging of lymphedema based on ICG lymphography, the classification proposed by Yamamoto et al was used: normal lymphatic pattern (stage 0; Video A, available online only); abnormal findings, such as splash and stardust patterns, in the inguinal

region only (stage 1); a stardust pattern above the knee, but normal lymph flow below the knee (stage 2); a stardust pattern in the inguinal region over the ankle, but normal lymph flow on the distal side from the ankle (stage 3); a stardust pattern spread to the toes (stage 4); and a diffuse pattern (stage 5; Video B, available online only).

Yoon et al. 2020

ICG lymphography images were taken by using a near- infrared camera to detect 760 nm light-emitting diode and a filter-cutting light below 820 nm. A total of 0.2 mL of ICG (Diagnogreen 2.5 mg/mL) was injected subcutaneously into the bilateral upper extremities at the first and third web space of the hand. To minimize the pain, lidocaine HCL 2% with epinephrine 1:100,000 was injected before dye injection. After injection, four aspects (ant, post, medial and lateral) of images of lymphatic drainage were obtained using the near-infrared camera. Images were taken every hour after injection until the ICG dye spread to the whole arm and sufficient ICG lymphography staging was possible. The images were classified into Type I - V according to arm dermal backflow (ADB) stage and modified MD Anderson Cancer Center (MDACC) stage based on the dermal back flow pattern and how it was extended.