Brentuximab vedotin (BV), an anti-CD30 antibody conjugated with the cytotoxic agent monomethyl auristatin E (MMAE), has achieved significant therapeutic success in CD30-positive cutaneous T-cell lymphomas (CTCL).1 Microscopic evaluation of immunohistochemically stained slides of tumoral tissue is regarded as the gold standard to select appropriate patients for CD30-targeted therapy.2 Among the different subgroups of CTCL, there is high variability of CD30 expression.3 Surprisingly, in mycosis fungoides (MF), which is the most common CTCL, profound treatment response to BV has also been observed when CD30 expression, as assessed by immunohistochemistry, was quite negligible (<5%, ie, expression levels that are commonly stated by pathologists as being negative).4, 5 This hitherto incompletely understood phenomenon—which might be related to insufficient sensitivity of immunohistochemical CD30 detection—is, however, of immediate clinical relevance: based on the approval conditions of BV, such “CD30-negative” patients might not be considered eligible for treatment with BV: that is, these patients would not receive one of the most efficient treatment options currently available for advanced CTCL.

Motivated by these observations, we aimed to evaluate whether the detection of CD30 mRNA expression by quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis would be more sensitive than immunohistochemistry when assessing the CD30 expression status of tumoral tissue.

A total of 18 samples (cryo-preserved, and formalin-fixed, paraffin-embedded tissue) obtained from skin lesions of seven patients with MF being treated with BV were analyzed (Table 1). The study was approved by the ethical committee of the University of Würzburg, and all patients provided informed consent. Biopsy samples were obtained before, during, and after BV treatment. The percentage of CD30 expression levels as assessed by immunohistochemistry using the anti-CD30 antibody Ber-H2 (Dako, Agilent Technologies, Santa Clara, California) was categorized into four different groups based on the frequency of CD30-positive cells (%) (Score 0, <5%; Score 1, 5%–25%; Score 2, 26%–50%; Score 3, 51%–75%; Score 4, >75%) (Figure 1A,B). In addition, staining intensity was assessed (faint/medium/high). qRT-PCR of mRNA isolated from corresponding cryo-preserved tissue sections was performed using TaqMan assay. For CD30 mRNA detection, the following specific CD30 primers and Taqman probe were used: CD30_fw: GAA TTC GGC AGA AGC TCC AC; CD30_rv: CTC CTC CTG GGT CTG GAA TC; CD30_probe: [6FAM]CCG GTC CAG ACC TCC CAG CC[BHQ1]. ACTB expression served as endogenous control (ß-actin_fw: GCG AGA AGA TGA CCC AGA TC; ß-actin_rv: CCA GTG GTA CGG CCA GAG G; ß-actin_probe: [6FAM]CCA GCC ATG TAC GTT GCT ATC CAG GC[BHQ1]). The relative CD30 mRNA expression was calculated by the ΔΔCt method (Sample #5 served as calibrator) and was compared with the immunohistochemical data (Figure 1B).

TABLE 1. CD30 expression in mycosis fungoides Patient number # Sample number # CD30 expression (IHC) CD30 score (IHC) CD30 expression intensity Relative CD30 expression (RT-PCR) Lesion type Timepoint of biopsy in relation to BV therapy BV therapy duration Response of biopsied lesion Best global response to BV therapy 1 1 99% 4 High 0.241 Tumor with LCT Before January 2019 to July 2019 CR 2 99% 4 High 0.243 Tumor with LCT After NAP MR 2 3 1% 0 Medium 0.042 Tumor with LCT During August 2019 to April 2020 PR 4 0% 0 Medium 0.047 Tumor with LCT During PD PR 3 5 99% 4 High 1 Tumor with LCT Before June 2020 to August 2020 CR 6 40% 2 Mediu-high 0.011 Tumor with LCT Before CR CR 4 7 5% 1 Medium-high 0.051 Plaque During January 2019 to May 2020 CR 8 2% 0 Medium 0.521 Tumor After NAP PR 9 5% 1 Medium 0.219 Tumor After NAP 5 10 2% 0 Medium 0.018 Tumor with LCT Before December 2019 to December 2020 CR 11 90% 4 High 0.146 Tumor with LCT Before CR PR 12 10% 1 Medium 0.272 Plaque Before PR 6 13 2% 0 Medium 0.107 Tumor Before May 2019 to August 2020 CR 14 15% 1 Medium 0.206 Plaque Before CR CR 15 15% 1 High 0.072 Plaque During CR 7 16 <1% 0 Medium 0.004 Tumor with LCT Before March 2019 to June 2019 NAP 17 4% 0 Medium 0.005 Tumor with LCT During PD PR 18 <<1% (3 positive cells) 0 Medium 0.035 Tumor with LCT During PD NC 19 <<1% (3 positive cells) 0 Faint 0 Inflammatory infiltrate NAP NAP NAP PC 20 99% 4 High 0.595 cALCL NAP NAP NAP Note: Clinical findings of MF patients with individual response to BV treatment in relation to CD30 expression. Abbreviations: BV, brentuximab vedotin; cALCL, cutaneous anaplastic large T-cell lymphoma; CR, complete response; IHC, immunohistochemistry; LCT, large cell transformation; MF, mycosis fungoides; MR, mixed response; na, not applicable; NC, negative control; PC, positive control; PR, partial response.

(A) Representative histopathological pictures (H&E, CD30 staining; magnification ×400) of skin biopsies of Patient #5 with low CD30 expression (Sample #10) (upper part) and high CD30 expression (Sample #11) (lower part). (B) Relative CD30 mRNA expression as calculated by the ΔΔCt method plotted against the results of semiquantitative immunohistochemical evaluation in 18 samples of MF patients treated with BV (Spearman's correlation coefficient r = 0.63). A sample of a cALCL (red triangle) served as positive control and an inflammatory infiltrate as negative control (blue square). BV, brentuximab vedotin; cALCL, cutaneous anaplastic large T-cell lymphoma; MF, mycosis fungoides; NC, negative control; PC, positive control

CD30 expression at the mRNA and protein level—as investigated by the two different methods (semiquantitative immunohistochemical assessment and qRT-PCR analysis)—showed a moderate correlation in the analyzed samples (Spearman's correlation coefficient r = 0.63). For MF samples from patients without or with only negligible CD30 expression on immunohistochemistry (≤1%, ie, only scattered single cells within the infiltrate) (n = 4), qRT-PCR analysis revealed very low levels of CD30 mRNA in all samples. Of note, the biopsied lesions demonstrated response to BV treatment rather independently of CD30 expression levels and even in samples with very low CD30 expression (eg, Samples #7, #10, or #13).

Although essential for a meaningful selection of appropriate patients for CD30-directed therapy, up to now there are no consensus criteria on standardized evaluation of CD30 immunolabeling of tumor tissue.6 Recently, by the means of digitally enhanced imaging techniques, CD30 expression could be indeed deciphered in >90% of immunohistochemically stained slides having previously been designated as CD30-negative by light microscopic evaluation.4 This suggests that in such cases the target molecule CD30 might nevertheless be expressed at the protein level but only at minor amounts below the detection threshold of visual evaluation. Although limited by the small number of patients, our data indicate that qRT-PCR analysis for CD30 detection in tumoral tissue does not contribute to a treatment-relevant categorization of patients with MF with regard to CD30-targeted therapy. Rather, regardless of CD30 expression status (as evidenced by immunohistochemistry or qRT-PCR quantification), patients with MF may respond to BV treatment.

The reason why BV shows significant clinical efficacy in MF even in CD30-negative/low cases7 remains, however, currently unknown. This fundamental open question could not be answered by this project, but will be the subject of future investigations. Knowledge of the intracellular and extracellular distribution of the target molecule is essential to thoroughly evaluate immunohistochemical stainings.8 Moreover, one might speculate that nonspecific cytotoxic or immunomodulatory properties of unbound MMAE on tumor and bystander cells might play a central role in such cases irrespective of CD30 expression.

Although similar investigations comparing qRT-PCR and immunohistochemistry with regard to CD30 expression are lacking in CTCL, previous investigations have addressed this issue with comparable results in peripheral T-cell lymphomas9 and angioimmunoblastic T-cell lymphomas.10

Taking together, our data show that patients without significant CD30 expression in biopsy samples, as evidenced by different assessment tools, might nevertheless profit from CD30-targeted treatment approaches. Validation and evaluation in larger patient cohorts might provide deeper insight into the biological functions of CD30 in MF and CD30-targeted treatment approaches.

ACKNOWLEDGMENT

The project was funded by the Interdisciplinary Center for Clinical Research (IZKF) of the University Hospital Würzburg (project B-411).

CONFLICT OF INTEREST

Marion Wobser received speaker's honoraria from TAKEDA. The other authors declare no potential conflict of interest.