Authors’ response

Dear Sir,

We thank Sharma et al1 for showing interest in our review2 and raising some queries. We wrote this article purely from an academic point of view without any conflicts of interest. We have not commented on a particular brand of medicine or manufacturer. As a pharmacologist working in academia, it is our moral and scientific obligation to critically appraise the published literature on varied aspects of medicines and their use in the community. Expressing views based on the published literature and raising concerns are well within our purview.

We have noticed that Sharma et al1 have made many vague, personal and derogatory remarks in their correspondence, which is undesirable and uncalled for. In many instances, it looks like that Sharma et al1 want to impose certain statements, which is totally unwarranted. Following is our detailed reply to the concerns raised by Sharma et al1.

They mentioned that Sharma et al2 failed to distinguish appropriately between different indications of tafenoquine (TQ) (radical cure of P. vivax malaria vs. malaria prophylaxis) and dosing regimens. We would like to state that our article covers the dosing regimen and duration of treatment in both the safety and efficacy part of the studies, for which we have provided and documented appropriate references. In case any reader wants to explore more knowledge on the topic, he/she can refer to the cited articles. Using terms such as inaccuracies and inconsistencies without any firm basis is not desirable; hence, should be avoided. This is a review article2, and we have commented based on data from the published articles and on the theme of our review. In the 1^st article cited (Walsh et al3, 1999) in the efficacy part of the review, we have clearly mentioned that this study was investigating the prevention of relapse of P. vivax malaria. Hence, there is no ambiguity. In the 2^nd article cited in our review in the efficacy part by Walsh et al4, it is clear that this study was looking at the protective efficacy of TQ in preventing relapse of P. vivax malaria. In the 3^nd article cited in our review in the efficacy part by Walsh et al5, it is very much evident from the results provided in terms of the protective efficacy of TQ that this study was assessing the efficacy of TQ in the prophylaxis of malaria. In the 4^th article cited (Elmes et al6), we have clearly mentioned that this study assessed the efficacy of three different doses of TQ for three days in P. vivax malaria as post-exposure prophylaxis vs. PQ plus doxycycline. Similarly, there is no confusion regarding the objective of the studies cited by us, Llanos-Cuentas et al7, Llanos-Cuentas8, Lacerda et al9, Nasveld et al10 and Fukuda et al11.

Another concern expressed by Sharma et al1 was regarding the intention of our review. We would like to draw the attention that the theme of our article2 was to discuss the potential concerns with respect to the efficacy and safety of TQ and the study designs of the published research articles. We agree with Sharma et al1 about the doses of TQ for the radical cure of P. vivax malaria and for malaria prophylaxis. Herein, we want to clarify that we did not specifically look into prophylaxis or radical cure; however, we have highlighted its efficacy and safety based on the retrieved published literature, which were accessible. It is surprising to see that Sharma et al1 are only interested in radical cure, which we never mentioned as a separate theme of our article2.

Other concerns expressed were regarding the article2 inaccurately outlining efficacy data from safety studies, confusing TQ dosages, inappropriately including certain studies, misinterpreting study results and overstating safety concerns. It is pertinent to mention here that we clearly wrote about the TQ doses, regimens and duration of treatment in the section ‘Efficacy of TQ observed in various studies’ part of our article2. There is no point of confusion as the doses and duration are clearly mentioned. There is no misinterpretation of the results, and we refute this statement. It seems that Sharma et al1 are unnecessarily highlighting the issue of prophylactic and radical cure. Regarding the safety aspects, we clearly mentioned in the remarks column of the table regarding the doses, duration and regimen of TQ.

As shown in the study done by Elmes et al6, TQ 200 mg once daily (approved dose as mentioned by you) was for three days. The extent of diarrhoea with TQ was 4.7 vs. 2.2 per cent in the PQ group. The extent of abdominal distress was 6.4 per cent in the TQ once daily group and 3.2 in the PQ group. The extent of diarrhoea with TQ 200 mg twice daily dosing was 14.9 per cent, whereas with primaquine (PQ), it was 2.2 per cent. The extent of abdominal distress was 10.6 in the TQ group vs. 3.2 per cent in the PQ group. Kindly refer to Table no. 2, page no 1099 in the article published by Elmes et al6.

Regarding the concern about discussing the phase 3 GATHER study and non-inferiority conclusion, we strongly rebut the above comment as we have clearly written this statement on page number 798, the second last paragraph of our article2, ‘In another randomized controlled trial study conducted by Llanos-Cuentas, analysis of phase 3 data regarding efficacy and safety outcomes was done along with patient-level meta-analysis ………. In the efficacy part, it was reported that 67 and 72.8 per cent of patients were free from recurrence at six months in the TQ and PQ groups, respectively’. Hence, the reccurrence-free rates were higher in the PQ group, and we highlighted the same (can be cross-checked with the published article). They (authors of the particular article) concluded that TQ failed to show non-inferiority compared to PQ’, and it was clearly stated in the conclusion of the cited article8.

Another concern was about ‘similar results’ from Fukuda et al11 and to compare the rates of thrombocytopenia, methemoglobinemia and for TQ vs. PQ. We used the term ‘similar results’11 to point out the trend of the results and not the doses of TQ. The dose of TQ used, i.e. 400 mg, has been clearly mentioned. Further, we want to draw your attention to an article published by Lacerda et al9, in 2019, in which the authors used a single approved dose of TQ. Sharma et al1 may kindly refer to table number 2, page number 224 and carefully compare the adverse events profile between TQ and PQ (Lacerda et al9, 2019). Moreover, a study done by Walsh et al4 in 2004 reported that TQ at a dose of 300 mg once daily for seven days developed methemoglobinemia in 12 per cent in the 300 mg group and 3.3 per cent in PQ 15 mg once daily for 14 days (page number 1099).

Regarding the concern about our conclusions of the DETECTIVE study7, it is pertinent to mention that as pharmacologists, we strongly feel it is not rational to compare six months’ relapse free efficacy rate for TQ in combination with CQ vs. CQ alone. We wish to impress on the learned authors that relapse in P. vivax malaria cases is due to the presence of hypnozoites in the liver, and the standard of care for preventing relapse is PQ to be given for 14 days, and CQ (effective as erythrocytic schizonticide) is not the primary therapy for preventing relapse, and a majority of the pharmacologists and infectious disease specialists will endorse this statement. Our statement is well endorsed by the malaria treatment guidelines as well12,13. Hence, we raised a concern regarding the study design of this particular article. In our view, this study should have compared TQ + CQ vs. PQ + CQ.

Regarding potential safety concerns such as keratopathy and retinal disorders, we discussed the potential safety concern of TQ. Our intention was not to underestimate or criticize this drug. We opined on the drug-related aspects. We clearly mentioned the word ‘higher dose’ of TQ rather than TQ. Kindly refer to page number 799, under potential safety concerns of TQ1. Regarding the study quoted by Sharma et al1, i.e. Ackert et al14, in 2019, we would like to inform you that this study was not available to us at the time of compiling our manuscript and that was the probable reason due to which we were not able to include this study as well as any studies or updates published after the submission of our manuscript. Furthermore, we fail to understand how discussing ‘TQ use in the prophylaxis setting’ can be misleading when one of the approved indications of TQ is prophylaxis of P. vivax malaria.

Regarding the concern about our statement that ‘it does not seem to be meaningful’ in ‘enroling CQ alone group in phase 2b component of the DETECTIVE study7, we agree that no study can be done without ethical approval. However, as pharmacologists, we hold the opinion that depriving these subjects from the standard of care for radical cure of P. vivax malaria (PQ) does not seem to be meaningful as relapse in P. vivax malaria cases is due to the presence of hypnozoites and standard of care for preventing relapse is PQ to be given for 14 days, and CQ (an erythrocytic schizonticide) is not the primary therapy for preventing relapse.

Sharma et al1 also raised the concern regarding the incidence of diarrhoea for TQ vs. PQ (16 vs. 8%). A study done by Lacerda et al9 showed that the extent of diarrhoea with SD TQ was 3.8 per cent, whereas with PQ it was 2.3 per cent. Although the primary objective of the study was not to compare TQ with PQ, it should have been considered in the final statistical analysis (see Table no. 2, page no 224 of the above article). Elmes et al6 used 200 mg once daily dose of TQ for three days. In our opinion, the authors should have compared that cohort with the PQ group. The extent of diarrhoea with TQ was 4.7 vs. 2.2 per cent in the PQ group. The extent of abdominal distress was 6.4 per cent in the TQ once-daily group and 3.2 in the PQ group. The extent of diarrhoea with TQ 200 mg twice daily was 14.9 per cent whereas in PQ was 2.2 per cent. The extent of abdominal distress was 10.6 in the TQ group vs. 3.2 per cent in the PQ group. Refer to Table no. 2, page no 1099 of the article published by Elmes et al6. We mentioned everything clearly in the table of adverse events of TQ and doses, dose regimen and duration of treatment.

It is pertinent to mention that ours was a review article based on the literature published on the stated topic and it was not a systematic review. It could be likely that we might have missed the USFDA NDA document, since we focused mainly on the published research articles based on the theme of our review.

Another concern was raised regarding the decline in the haemoglobin level with TQ. Notably, the study done by Lacerda et al9 in 2019 showed that a decline in the haemoglobin level with TQ was found to be 5.4 per cent compared to 1.6 per cent in the PQ arm, which was more than three times higher with TQ (page no 223 and 224 of the article published by Lacerda et al9). It was clearly mentioned in the above study that in the TQ group, seven subjects met the protocol-defined haematological stopping criteria. We used the terms like potential safety concerns because we highlighted the potential concerns, which needed attention, and further data might be required. We refute the repeated use of words like ‘emotive’, ‘unqualified’, ‘misinterpretation’ and ‘overstated’ since we have given adequate references and explanations.

For reasons, best known to the learned authors, they completely overlooked the paragraph on page number 803 of our article, which clearly mentioned the potential benefits of TQ over PQ2:‘In a developing country like India, TQ can be…. can increase the adherence rates dramatically’. Similarly, the authors1 completely overlooked the lines indicating that TQ showed promising results (1^st paragraph of page 804)2.

In the concluding remarks of our article, we clearly mentioned the edge of TQ over PQ in terms of adherence, and we wanted to consider the potential safety concerns, which hold true for any new drug. Furthermore, we highlighted the need for future research and implementation of certain mitigation strategies to address the concerns related to the safety and efficacy. The concern regarding the conclusion of the paper reporting pooled GATHER and DETECTIVE data was unfounded as in this particular part of our review, we talked about DETECTIVE trial7 suggesting that the statistical comparison should have been done for the safety part as well; we did not talk about pooled analysis.

Another concern expressed was regarding failure to acknowledge studies 2012/03/002511 (Clinical Trials Registry-India) and NCT01376167 (ClinicalTrials.gov). Notably, in this particular part of our article2, we wanted to highlight the current status of TQ with respect to ongoing research and regulatory approval process, and we came across two different registration numbers, one with ClinicalTrials.gov (NCT01376167) and another with the Clinical Trials Registry of India (CTRI, CTRI/2012/03/002511). We do not feel that providing both these references (indicating the international and national level ongoing trial activities) could mislead the readers, and it has nothing to do with the data generated subsequently. The CTRI registration is a dynamic process, and there is a provision to update the status of the trial registered with CTRI. It is very much likely that when we accessed this trial on the CTRI website (on July 4, 2019, as given in reference number 23 of our published article2), we cited the data accordingly, and later on, the status could have been changed. We carefully and correctly mentioned all the doses as given in these registries. Hence, we refute these comments.

Regarding the concern about the risk of inadvertent use of TQ in G6PD-deficient subjects, we suggest to refer carefully to our full statement (page 804, paragraph 3)2. This statement was issued in the Recommendations of the SEC (Antimicrobial and Antiviral) made in its 56^th meeting held on April 25, 2019 at CDSCO, HQ (reference 25 in our article2), as follows: ‘Tafenoquine has a disadvantage in case use of inadvertent use of the drug in G6PD deficient patients, as with single dose administration, the effect of the drug will remain in the body for considerable period of days’. Hence, the above concern was not raised by us; however, we just highlighted it by citing the proper reference.

To conclude, it must be stated that the correspondence by Sharma et al1 has not raised any significant and valid concern which requires any correction. Rather, it was unnecessarily repetitive, vague, exhaustive and full of unsavoury remarks. As science is ever-evolving based on evidence generated through ongoing research, academicians and researchers can present their views based on the current evidence.