direct immunofluorescence of skin using formalin-fixed ... · of the formalin-fixed...

J Clin Pathol 1980; 33: 365-369

Direct immunofluorescence of skin usingformalin-fixed paraffin-embedded sectionsSL MERA, EW YOUNG, AND JWB BRADFIELD

From the University Department ofPathology, University Walk and Bristol Royal Infirmary, Bristol,Avon, UK

SUMMARY The technique of direct immunofluorescence has been applied to skin biopsy specimensfixed in formalin and embedded in paraffin wax. The results have been compared with those obtainedby using snap-frozen biopsy specimens from the same patients. Trypsinisation of the dewaxedmaterial allowed subsequent detection of immunoglobulins, complement, and fibrinogen. Whencompared to the fluorescence in the snap-frozen specimens, the staining in the paraffin sections wasless bright and there was a higher rate of negatives. Even so, it was possible to establish the diag-nosis in most cases of pemphigus, pemphigoid, and lupus erythematosus.

The technique of direct immunofluorescence todemonstrate immunoglobulins, complement, andfibrinogen is of proven value in the diagnosis ofseveral different skin conditions.l 2 One of the majordisadvantages of this procedure is the requirementfor fresh frozen tissue. Problems arise because (1)facilities for the snap-freezing of fresh specimens arenot always available in all clinical areas, (2) thetransportation of frozen tissue may be difficult,especially between hospitals, and the use of specialholding fixatives to allow transportation beforefreezing may lead to loss of antigen reactivity,3 (3)the long-term storage of frozen tissue also leads toloss of antigen reactivity.4 Finally, formalin-fixedmaterial is sometimes the only tissue availableeither because the lesion was too small to divide orbecause stored material is being reviewed retro-spectively. All of these disadvantages would beavoided if formalin-fixed paraffin-embedded tissuescould be used for direct immunofluorescence. Wedescribe here the results of a study which was doneto assess the possibility of using such material in adiagnostic skin immunofluorescence service.

Material and methods

SELECTION OF BIOPSY MATERIALSix cases each of pemphigus, pemphigoid, and lupuserythematosus were studied. The diagnosis wasestablished using clinical features, conventionallight microscopy, and direct immunofluorescence of

Received for publication 4 October 1979

fresh frozen tissue. The six control biopsies were frompatients with unrelated skin disorders whichshowed minimal histological abnormalities andentirely negative direct immunofluorescence onfrozen tissue.

PREPARATION OF BIOPSY TISSUETwo 5 mm punch biopsies were obtained from eachpatient.

Cryostat sectionsOne of the two specimens was immediately snap-frozen in liquid nitrogen. After storage in liquidnitrogen for not longer than seven days the frozentissue was mounted in OCT compound (Lab-Tek,Ames) and sectioned at 6)u in a cryostat at - 20'C.The sections were then stained by the direct immuno-fluorescence procedure without prior fixation ortrypsinisation.

Paraffin sectionsThe other punch biopsy was fixed in 100% non-buffered formol saline, processed, and embedded inparaffin wax using standard laboratory procedures.Five micron wax sections were cut. For histologicalstaining, these were dewaxed and stained withhaematoxylin and eosin. For immunofluorescence,sections were dried overnight at 370C.

TRYPSINISATIONBefore immunofluorescent staining, dewaxed andrehydrated sections were rinsed in Tris-saline buffer,pH 7-8, and then trypsinised at 370C in a solution of

365

on 31 March 2019 by guest. P

rotected by copyright.http://jcp.bm

j.com/

J Clin P

athol: first published as 10.1136/jcp.33.4.365 on 1 April 1980. D

ownloaded from

http://jcp.bmj.com/

Mera, Young, and Bradfield

Tris-saline buffer containing 0-1 % calcium chlorideand 005% trypsin (Wellcome). After a series ofpreliminary investigations to document the effects ofvarying the times of trypsinisation (see Resultssection) a standard incubation time of 40 minuteswas adopted. After trypsin treatment the sectionswere rinsed and left overnight at 40C in Tris-salinebuffer.

IMMUNOFLUORESCENT STAININGBoth frozen and trypsinised formalin-fixed sectionswere treated similarly. Before staining they wererinsed with two changes of phosphate bufferedsaline (PBS), pH 7-3. They were then incubated in amoist chamber for 30 minutes at room temperaturewith commercially available fluorescein conjugatesof rabbit-antihuman IgG, IgM, IgA, C3, andfibrinogen (Dakopatts, Mercia Brocades). Afterstaining they were washed in three changes of PBSbefore being mounted in PBS/glycerol.The effect of variations in the titres of antisera

were initially investigated using serial dilutionsranging from neat antisera to 1/320. There wasoptimal staining when the immunofluorescence wasat a maximum compared with the backgroundfluorescence; this occurred at a dilution of 1/10 foreach of the conjugated antisera in both frozen andparaffin sections. It was not possible to raise theintensity of fluorescence in the fixed material byincreasing the concentration of the reagents to morethan 1/10 without causing an unacceptable level ofbackground fluorescence. A dilution of 1/10 wastherefore used for the staining of both frozen andparaffin sections.

SPECIFICITYThe purpose of this study was to compare the pat-terns of skin immunofluorescence in frozen andformalin-fixed paraffin-embedded sections from thesame patients. Because of this the possibility ofvariation in staining patterns between individualbatches of antisera was excluded not by specificblocking procedures but by the incorporation ofknown positive and negative controls. In addition,all the paraffin-embedded material was stained on atleast two separate occasions with different batchesof antisera.

EXAMINATION OF TREATED SECTIONSAll the preparations were examined independentlyby at least two investigators using a Zeiss Universalmicroscope equipped with FITC excitor and barrierfilters and incident illumination from an Osram HBO50 lamp. Staining patterns were noted and theintensity of fluorescence was graded from 0 to+ + +.

Results

GENERALIn the formalin-fixed paraffin-embedded materialIgG, IgM, IgA, C3, and fibrinogen were all detected.Trypsin pretreatment was essential. A preliminarystudy showed that when trypsin incubation times ofless than 20 minutes were used the numbers and theintensity of positive results were both reduced. Incontrast, treatment for more than 1 hour led notonly to poorer discrimination between specificimmunofluorescence and background staining butalso to an unacceptably high loss of sections fromthe slides.The patterns of immunofluorescence were identical

in both fixed and frozen material, although theintensity of fluorescence was usually less in thesections of fixed tissue. In the control groups,selected because of entirely negative directimmunofluorescence on frozen section, there wereno positive results using formalin-fixed paraffin-embedded sections.

PEMPHIGUSWithin the epidermis intercellular staining for IgGcould be demonstrated in most of the formalin-fixedparaffin-embedded sections studied, and its pattern(Fig. 1) exactly resembled that seen in frozenmaterial. Table 1 shows that, of the six casesselected for their intercellular staining for IgG onfrozen section, five were positive for IgG usingparaffin-embedded material, although the intensityof fluorescence was usually reduced. The remainingone negative case remained so on repeat study. Thecases that were positive for IgM and IgA on frozensection were all negative on the paraffin-embeddedmaterial. The five cases that on frozen section showedpositive staining for C3, and the three that werepositive for fibrinogen, were all negative usingparaffin-embedded sections. None of the paraffinsections gave positive immunofluorescence unless thefrozen sections had also been positive.

PEMPHIGOIDImmunofluorescent staining for IgG at the base-ment membrane zone could be demonstrated in allof the formalin-fixed paraffin-embedded sectionsstudied, and its pattern, shown in Fig. 2, exactlyresembled that seen in the frozen material. Table 2shows that, of the six cases selected for their stainingfor IgG at the basement membrane zone in frozensections, all six showed similar staining usingparaffin-embedded sections. A similar pattern ofstaining for C3 was seen in paraffin-embeddedsections in four out of the five cases that werepositive in the frozen sections. In most cases the

366



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Direct immunofluorescence of skin using formalin-fixed paraffin-embedded sections

Fig. 1 Pemphigus: intercellularIgG. Paraffin section, directimmunofluorescence x 500

Table 1 Pemphigus: intercellular staining

Case Tissue IgG IgM IgA C3 Fibrin

1 Frozen +++ + + +++ ++Paraffin ++ - - - -

2 Frozen + + + + +++Paraffin + + -

3 Frozen + + - + + + + +Paraffin + -

4 Frozen + + - - +Paraffin + -

5 Frozen -{+ -

Paraffin - - - - -6 Frozen + - - +± + +

Paraffin +

fluorescence was less bright in the paraffin-embeddedmaterial. Of the two cases that showed positivestaining for IgM on frozen section, only one waspositive in the paraffin-embedded material. Of thetwo cases that were positive for IgA in frozensections, neither was positive in paraffin sections. Ofthe two that were positive for fibrinogen, one waspositive in the paraffin-embedded material. Inaddition, one case that stained for fibrinogen in theparaffin section had been negative in the frozentissue.

LUPUS ERYTHEMATOSUSImmunofluorescent staining for IgG, IgM, IgA, andfibrinogen at the dermal-epidermal junction wasseen in formalin-fixed paraffin-embedded material.The pattern was thick and granular (Fig. 3),similar to that seen in frozen sections. The fullresults are shown in Table 3. Of the six cases thatshowed positive staining for IgG and IgM in frozensections, one was negative for IgG, and another wasnegative for 1gM, using paraffin-embedded material.On staining for IgA, five of the frozen sections werepositive but only two of these were positive in theparaffin sections. Even though C3 was detected inall six cases in frozen section, all were negative in theparaffin-embedded material. Fibrinogen was demon-strated in the frozen sections in all six cases but onlyfour cases were positive using paraffin sections.

Discussion

This study shows that the technique of directimmunofluorescence can be applied successfully toformalin-fixed paraffin-embedded skin biopsy speci-mens. Previous reports of immunofluorescence

Fig. 2 Pemphigoid: basementmembrane zone IgG. Paraffinsection, direct immunofluor-escence x 500

367



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

368

Table 2 Pemphigoid: basement membrane zone staining

Case Tissue IgG 1gM IgA C3 Fibrin

7 Frozen + + + + + + + + + +Paraffin t +- ++ +

8 Frozen i-t + + -Paraffin -+- -

9 Frozen + + + + +Paraffin + + + + +

10 Frozen +- +-+ + +Paraffin + - -

11 Frozen + - ++ -

Paraffin + - - +-f -

12 Frozen - - - -Paraffin -t

using paraffin-embedded material have includedstudies of lymphoid, thyroid, renal, pulmonary, andintestinal tissues,5 6 and the demonstration ofintracellular antigens.6 7

In the present study dewaxed sections of theparaffin-embedded skin specimens were treated withtrypsin as this is known to improve the detection ofantigens.8 9 Even so, we have noted some loss ofsensitivity when compared to the snap-frozenmaterial in that the immunofluorescence in theparaffin-embedded sections was often diminished inintensity, and in some instances the antigens wereundetectable. Attempts to improve the sensitivity byincreasing the titre of antisera were unsuccessful.Some of the negative results, and the additionalpositive result for fibrinogen in the paraffin sectionsof case 9, may have been due to variation in thepresence of antigens between different sites of biopsy;certainly deposits were sometimes present only inone area within any given biopsy site.Of the antigens sought, C3 was the most difficult

to demonstrate in formalin-fixed paraffin-embeddedmaterial. We were unable to detect C3 in any of theparaffin-embedded biopsy specimens of pemphigus

Mera, Younig, and Bradfield

or lupus erythematosus. Since then we have detectedC3 in other cases of lupus erythematosus but not, sofar, in paraffin-embedded material from pemphigus.In contrast, it proved easy to demonstrate C3 incases of pemphigoid.Taken together, these results show that the

correct diagnosis would have been made in 17 out ofthe 18 cases examined if formalin-fixed paraffin-embedded sections had been the only tissue available.Conversely, there would have been no false-positivediagnoses. Some artifactual epidermal fluorescencewas occasionally encountered, but so far this has beeneasily recognised as such. In pemphigus, the immuno-fluorescence was entirely intercellular; in the upperepidermis it was characterised by a fine stainingpattern between the cells, and in the mid epidermisit showed a spotted or speckled pattern (Fig. 1). Incontrast, artifactual epidermal cell fluorescence wascharacterised by a patchy and denser fluorescence,which often involved the cells themselves, through allepidermal layers and where it was intercellular therewas complete absence of any spotted appearance.

In many recent studies using formalin-fixedparaffin-embedded material, immunoperoxidase

Table 3 Lupus erythematosus: dermal-epidermaljunction staining

Case Tissue IgG IgM IgA C3 Fibrin

13 Frozen + + +-+ tr+ +1-i-Paraffin + + + -- +

14 Frozen + + + + + + +Paraffin + + -+- + +

15 Frozen +- - ±Paraffin + + + +

16 Frozen + + +-m- + +Paraffin -

17 Frozen + +Paraffin - - -

18 Frozen + + -+Paraffin +-r - - +

Fig. 3 Lupus erythematosus:dermal-epidermal juction IgG.Paraffin section, direct immuno-fluorescence x 500



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

Direct immunofluorescence of skin using formalin-fixed paraffin-embedded sections 369

techniques have been used to detect immunedeposits.1011 However, when compared with theimmunofluorescence technique, immunoperoxidasetechniques are not widely used and have provedmore difficult to establish as a regular and repro-ducible diagnostic investigation. It is for this reasonthat immunofluorescence was chosen for evaluationin this study.

In summary, the present results show that it ispossible to use the technique of direct immuno-fluorescence to demonstrate immunoglobulins, com-plement, and fibrinogen in formalin-fixed paraffin-embedded tissue sections of skin. Although thetechnique is less sensitive than when using fiozenmaterial the rate of diagnostic misinterpretation islow. Thus, when frozen tissue is not available,paraffin-embedded material can be used for directimmunofluorescence in the diagnosis of skin disease.

We are grateful to the Special Trustees of theMedical Research Committee, Bristol HealthDistrict (Teaching) for financial support. We alsothank Drs Martlew and Ellis for allowing us accessto material from Princess Margaret Hospital,Swindon.

References

Jablonska S, Beutner EH, Michel B, et al. Uses forimmunofluorescence tests of skin and sera; utilizationof immunofluorescence in the diagnosis of bullousdiseases, lupus erythematosus, and certain otherdermatoses. Arch Dermatol 1975;111:371-81.

2 Tuffanelli DL. Cutaneous immunopathology: Recentobservations. J Invest Dermatol 1975 ;65 :143-53.

3 Skeete MVH, Black MM. The evaluation of a special

liquid fixative for direct immunofluorescence.Clin Exp Dermatol 1977 ;2 :49-56.

4Beutner EH, Hale WL, Nisengard RJ, Chorzelski TP,Holubar K. In Immunopathology of the Skin:Labeled Antibody Studies. EH Beutner,TP Chorzelski,SF Bean, RE Jordan, eds. Stroudsburg, Pennsylvania:Dowden, Hutchinson, and Ross, 1973:213-7.

5Dorsett BH, loachim HL. A method for the use ofimmunofluorescence on paraffin-embedded tissues.Am J Clin Pathol 1978;69:66-72.

6 Curran RC, Gregory J. Demonstration of immuno-globulin in cryostat and paraffin sections of humantonsil by immunofluorescence and immunoper-oxidase techniques. J Clin Pathol 1978;31 :974-83.

Burns J, Hambridge M, Taylor CR. Intracellularimmunoglobulins. A comparative study on threestandard tissue processing methods using horse-radish peroxidase and fluorochrome conjugates.J Clin Pathol 1974 ;27 :548-57.

8 Huang SN, Minassian H, More JD. Application ofimmunofluorescent staining on paraffin sectionsimproved by trypsin digestion. Lab Invest 1976;35:383-90.

9 Curran RC, Gregory J. The unmasking of antigens inparaffin sections of tissue by trypsin. Experientia1977 ;33 :1400-1.

10 Taylor CR, Burns J. The demonstration of plasmacells and other immunoglobulin-containing cellsin formalin-fixed paraffin-embedded tissues usingperoxidase-labelled antibody. J Clin Pathol 1973 ;27:14-20.

Burns J. Background staining and sensitivity of theunlabelled antibody-enzyme (PAP) method. Com-parison with the peroxidase labelled antibodysandwich method using formalin fixed paraffinembedded material. Histochemistry 1975;43:291-4.

Requests for reprints to: Dr JWB Bradfield, Depart-ment of Pathology, Medical School, University Walk,Bristol BS8 1TD, UK.



j.com/

J Clin P


ownloaded from

http://jcp.bmj.com/

direct immunofluorescence of skin using formalin-fixed ... · of the formalin-fixed...

Documents